Context Aware Computing For IoTs - Survey

414
IEEE COMMUNICATIONS SURVEYS & TUTORIALS, VOL. 16, NO. 1, FIRST QUARTER 2014
Context Aware Computing for

The Internet of Things: A Survey
Charith Perera, Student Member, IEEE, Arkady Zaslavsky, Member, IEEE, Peter Christen, and
Dimitrios Georgakopoulos, Member, IEEE
AbstractAs we are moving towards the Internet of Things
(IoT), the number of sensors deployed around the world is
growing at a rapid pace. Market research has shown a significant
growth of sensor deployments over the past decade and has
predicted a significant increment of the growth rate in the
future. These sensors continuously generate enormous amounts
of data. However, in order to add value to raw sensor data
we need to understand it. Collection, modelling, reasoning, and
distribution of context in relation to sensor data plays critical
role in this challenge. Context-aware computing has proven to
be successful in understanding sensor data. In this paper, we
survey context awareness from an IoT perspective. We present
the necessary background by introducing the IoT paradigm and
context-aware fundamentals at the beginning. Then we provide
an in-depth analysis of context life cycle. We evaluate a subset
of projects (50) which represent the majority of research and
commercial solutions proposed in the field of context-aware
computing conducted over the last decade (2001-2011) based
on our own taxonomy. Finally, based on our evaluation, we
highlight the lessons to be learnt from the past and some
possible directions for future research. The survey addresses
a broad range of techniques, methods, models, functionalities,
systems, applications, and middleware solutions related to context
awareness and IoT. Our goal is not only to analyse, compare
and consolidate past research work but also to appreciate their
findings and discuss their applicability towards the IoT.
Index TermsInternet of things, context awareness, sensor
networks, sensor data, context life cycle, context reasoning,
context modelling, ubiquitous, pervasive, mobile, middleware.
I. I NTRODUCTION
ONTEXT awareness, as a core feature of ubiquitous
and pervasive computing systems, has existed and been
employed since the early 1990s. The focus on context-aware
computing evolved from desktop applications, web applications, mobile computing, pervasive/ubiquitous computing to
the Internet of Things (IoT) over the last decade. However,
context-aware computing became more popular with the introduction of the term ubiquitous computing by Mark Weiser
[1] in his ground-breaking paper The Computer for the 21st
Century in 1991. Then the term context-aware was first used
by Schilit and Theimer [2] in 1994.
Since then, research into context-awareness has been established as a well known research area in computer science.
Many researchers have proposed definitions and explanations
Manuscript received September 28, 2012; revised January 30, 2013.

C. Perera, A. Zaslavsky and D. Georgakopoulos are with the Information and Communication Centre, Commonwealth Scientific and Industrial Research Organisation, Canberra, ACT, 2601, Australia (e-mail: firstname.lastname@csiro.au).
P. Christen is with the Research School of Computer Science, The
Australian National University, Canberra, ACT 0200, Australia (e-mail:
peter.christen@anu.edu.au).
Digital Object Identifier 10.1109/SURV.2013.042313.00197
of different aspects of context-aware computing, as we will

discuss briefly in Section III. The definitions for context and
context-awareness that are widely accepted by the research
community today were proposed by Abowd et al. [3] in 1999.
During the last two decades, researchers and engineers
have developed a significant amount of prototypes, systems,
and solutions using context-aware computing techniques. Even
though the focus varied depending on each project, one aspect
remained fairly unchanged: that is the number of data sources
(e.g. software and hardware sources). For example, most of
the proposed solutions collect data from a limited number of
physical (hardware) and virtual (software) sensors. In these
situations, collecting and analysing sensor data from all the
sources is possible and feasible due to limited numbers. In
contrast, IoT envisions an era where billions of sensors are
connected to the Internet, which means it is not feasible to
process all the data collected by those sensors. Therefore,
context-awareness will play a critical role in deciding what
data needs to be processed and much more.
Due to advances in sensor technology, sensors are getting more powerful, cheaper and smaller in size, which has
stimulated large scale deployments. As a result, today we
have a large number of sensors already deployed and it is
predicted that the numbers will grow rapidly over the next
decade [4]. Ultimately, these sensors will generate big data
[5]. The data we collect may not have any value unless we
analyse, interpret, and understand it. Context-aware computing
has played an important role in tackling this challenge in
previous paradigms, such as mobile and pervasive, which lead
us to believe that it would continue to be successful in the
IoT paradigm as well. Context-aware computing allows us
to store context1 information linked to sensor data so the
interpretation can be done easily and more meaningfully. In
addition, understanding context makes it easier to perform
machine to machine communication as it is a core element
in the IoT vision.
When large numbers of sensors are deployed, and start
generating data, the traditional application based approach (i.e.
connect sensors directly to applications individually and manually) becomes infeasible. In order to address this inefficiency,
significant amounts of middleware solutions are introduced by
researchers. Each middleware solution focuses on different aspects in the IoT, such as device management, interoperability,
1 The term context implicitly provide the meaning of information according to the widely accepted definition provided by [3]. Therefore, it
is inaccurate to use the term context information where information is
explicitly mentioned. However, research community and documents on the
web frequently use the term context information. Therefore, we also use
both terms interchangeably.
c 2014 IEEE
1553-877X/14/$31.00
PERERA et al.: CONTEXT AWARE COMPUTING FOR THE INTERNET OF THINGS: A SURVEY
platform portability, context-awareness, security and privacy,

and many more. Even though, some solutions address multiple
aspects, an ideal middleware solution that addresses all the
aspects required by the IoT is yet to be designed. In this
survey, we consider identifying the context-aware computing
related features and functionalities that are required by an ideal
IoT middleware solution as a key task.
There have been several surveys conducted in relation to
this field. We briefly introduce these surveys in chronological
order. Chen and Kotz [6] (2000) have surveyed context awareness, focusing on applications, what context they use, and
how contextual information is leveraged. In 2004, Strang and
Linnhoff-Popien [7] compared the most popular context modelling techniques in the field. Middleware solutions for sensor
networks are surveyed by Molla and Ahamed [8] in 2006. Two
separate surveys were conducted by Kjaer [9] and Baldauf et
al. [10] in 2007 on context-aware systems and middleware
solutions using different taxonomies. Both surveys compared
limited numbers, but different projects with very little overlap.
c et al. [11] (2009) reviewed popular context representation
and reasoning from a pervasive computing perspective. In
2010, Bettini et al. [12] also comprehensively surveyed context
modelling and reasoning by focusing on techniques rather
than projects. In the same year another survey was done
by Saeed and Waheed [13] focusing on architectures in the
context-aware middleware domain. Bandyopadhyay et al. [14]
have conducted a survey on existing popular Internet of
Things middleware solutions in 2011. In 2012, Makris et al.
[15] have conducted a survey on context-aware mobile and
wireless networking (CAMoWiN) domain where they have
identified all the possible components of a typical CAMoWiN
architecture. The latest survey is done by Bellavista et al. [16]
(2013) which is focused on context distribution for mobile
ubiquitous systems.
Our survey differs from the previous literature surveys
mentioned above in many ways. Most of the surveys evaluated
a limited number of projects. In contrast, we selected a large
number of projects (50) covering a decade, based on the
unique criteria that will be explained at the end of this section.
These projects are different in scale. Some are large scale
projects and others corresponds to small scale contributions.
We took a much broader viewpoint compared to some of the
previous surveys, as they have focused on specific elements
such as modelling, reasoning, etc. Finally and most importantly, our taxonomy formation and organisation is completely
different. Rather than building a theoretical taxonomy and
then trying to classify existing research projects, prototypes
and systems according to it, we use a practical approach.
We built our taxonomy based on past research projects by
identifying the features, models, techniques, functionalities
and approaches they employed at higher levels (e.g. we do not
consider implementation/code level differences between different solutions). We consolidated this information and analysed
the capabilities of each solution or the project. We believe this
approach allows us to highlight the areas where researchers
have mostly (priorities) and rarely (non-priorities) focused
their attention and the reasons behind. Further, we have also
used a non-taxonomical project based evaluation, where we
highlight how the different combinations of components are
415
designed, developed and used in each project. This allows to

discuss their applicability from an IoT perspective.
Our objectives in revisiting the literature are threefold:
1) to learn how context-aware computing techniques have
helped to develop solutions in the past, 2) how can we apply
those techniques to solve problems in the future in different
paradigms such as the IoT, and 3) to highlight open challenges
and to discuss future research directions.
This paper is organised into sections as follows: Section
II provides an introduction to the IoT. In this section, we
briefly describe the history and evolution of the Internet.
Then we explain what the IoT is, followed by a list of
application domains and statistics that show the significance
of the IoT. We also describe the relationship between sensor
networks and the IoT. Comparisons of popular IoT middleware
solutions are presented at the end of the section in order to
highlight existing research gaps. In Section III, we present
context awareness fundamentals such as context-aware related
definitions, context types and categorisation schemes, features
and characteristics, and context awareness management design
principles. In Section IV, we conduct our main discussion
based on context life cycle where we identify four stages:
acquisition, modelling, reasoning, and distribution. Section V
briefly discusses the highlights of each project, which we use
for the comparison later. Finally, Section VI discusses the
lessons learn from the literature and Section VII identifies
future research directions and challenges. Conclusion remarks
are presented in Section VIII.
For this literature review, we analyse, compare, classify a
subset of both small scale and large scale projects (50) which
represent the majority of research and commercial solutions
proposed in the field of context-aware computing based on
our own taxonomy. We selected the existing solutions to
be reviewed based on different criteria. Mainly, we selected
projects that were conducted over the last decade (2001-2011).
We also considered main focus, techniques used, popularity,
comprehensiveness, information availability, and the year of
publication, in order to make sure that our review provides a
balanced view on context-aware computing research.
II. T HE I NTERNET OF T HINGS PARADIGM
In this section, we briefly introduce the IoT paradigm.
Our intention is not to survey the IoT, but to present some
fundamental information (e.g. how Internet evolved, what
is the IoT, statistics related to IoT, underline technologies,
characteristics, and research gaps in IoT paradigm) that will
help with understanding the historic movements and the
direction into which technology is moving today. The IoT
paradigm has its own concepts and characteristics. It also
shares significant amounts of concepts with other computer
fields. The IoT bundles different technologies (e.g. sensor
hardware/firmware, semantic, cloud, data modelling, storing,
reasoning, processing, communication technologies) together
to build its vision. We apply the existing technologies in
different ways based on the characteristics and demands of
the IoT. The IoT does not revolutionise our lives or the field
of computing. It is another step in the evolution of the Internet
we already have.
416

Fig. 1. Evolution of the Internet in five phases. The evolution of Internet begins with connecting two computers together and then moved towards creating
World Wide Web by connecting large number of computers together. The mobile-Internet emerged by connecting mobile devices to the Internet. Then, peoples
identities joined the Internet via social networks. Finally, it is moving towards Internet of Things by connecting every day objects to the Internet.
A. Evolution of Internet
Before we investigate the IoT in depth, it is worthwhile
to look at the evolution of the Internet. In the late 1960s,
communication between two computers was made possible
through a computer network [17]. In the early 1980s the
TCP/IP stack was introduced. Then, commercial use of the
Internet started in the late 1980s. Later, the World Wide Web
(WWW) became available in 1991 which made the Internet
more popular and stimulate the rapid growth. Web of Things
(WoT) [18], which based on WWW, is a part of IoT.
Later, mobile devices connected to the Internet and formed
the mobile-Internet [19]. With the emergence of social networking, users started to become connected together over the
Internet. The next step in the IoT is where objects around us
will be able to connect to each other (e.g. machine to machine)
and communicate via the Internet [20]. Figure 1 illustrates the
five phases in the evolution of the Internet.
B. What is the Internet of Things?
During the past decade, the IoT has gained significant
attention in academia as well as industry. The main reasons
behind this interest are the capabilities that the IoT [22], [23]
will offer. It promises to create a world where all the objects
(also called smart objects [24]) around us are connected to
the Internet and communicate with each other with minimum
human intervention [25]. The ultimate goal is to create a
better world for human beings, where objects around us know
what we like, what we want, and what we need and act
accordingly without explicit instructions [26].
The term Internet of Things was firstly coined by Kevin
Ashton [27] in a presentation in 1998. He has mentioned
The Internet of Things has the potential to change the
world, just as the Internet did. Maybe even more so. Then,
the MIT Auto-ID centre presented their IoT vision in 2001
[28]. Later, IoT was formally introduced by the International
Telecommunication Union (ITU) by the ITU Internet report
in 2005 [29].
The IoT encompasses a significant amount of technologies

that drive its vision. In the document, Vision and challenges
for realising the Internet of Things, by CERP-IoT [4], a
comprehensive set of technologies was listed. IoT is a very
broad vision. The research into the IoT is still in its infancy.
Therefore, there arent any standard definitions for IoT. The
following definitions were provided by different researchers.
Definition by [30]: Things have identities and virtual
personalities operating in smart spaces using intelligent
interfaces to connect and communicate within social, environment, and user contexts.
Definition by [20]:The semantic origin of the expression is
composed by two words and concepts: Internet and Thing,
where Internet can be defined as the world-wide network
of interconnected computer networks, based on a standard
communication protocol, the Internet suite (TCP/IP), while
Thing is an object not precisely identifiable Therefore, semantically, Internet of Things means a world-wide network
of interconnected objects uniquely addressable, based on
standard communication protocols.
Definition by [21]: The Internet of Things allows people
and things2 to be connected Anytime, Anyplace, with Anything and Anyone, ideally using Any path/network and Any
service.
We accept the last definition provided by [21] for our
research work, because we believe, this definition encapsulates
the broader vision of IoT. Figure 2 illustrates the definition
more clearly. The broadness of IoT can be identified by
evaluating the application domains presented in Section II-C.
C. IoT Application Domains
The IoT, interconnection and communication between everyday objects, enables many applications in many domains.
The application domain can be mainly divided in to three categories based on their focus [23], [4]: industry, environment,
2 We use both terms, objects and things interchangeably to give the same
meaning as they are frequently used in IoT related documentation. Some other
terms used by the research community are smart objects, devices, nodes.

Fig. 2. Definition of the Internet of Things: The Internet of Things allows

people and things to be connected anytime, anyplace, with anything and
anyone, ideally using any path/network and any service [21].
and society. The magnitude of the applications can be seen in

the statistics presented in Section II-D.
Supply chain management [31], transportation and logistics [32], aerospace, aviation, and automotive are some of
the industry focused applications of IoT. Telecommunication,
medical technology [33], healthcare, smart building, home
[34] and office, media, entertainment, and ticketing are some
of the society focused applications of IoT. Agriculture and
breeding [35], [36], recycling, disaster alerting, environmental
monitoring are some of the environment focused applications.
Asin and Gascon [37] listed 54 application domains under
twelve categories: smart cities, smart environment, smart water, smart metering, security and emergencies, retail, logistics,
industrial control, smart agriculture, smart animal farming,
domestic and home automation, and eHealth.
D. IoT Related Statistics
The vision of the IoT is heavily energised by statistics and
predictions. We present the statistics to justify our focus on
the IoT and to show the magnitude of the challenges. It is
estimated that there about 1.5 billion Internet-enabled PCs and
over 1 billion Internet-enabled mobile phones today. These
two categories will be joined with Internet-enabled devices
(smart objects [24])) in the future. By 2020, there will be 50
to 100 billion devices connected to the Internet [4].
According to BCC Research [38], the global market for
sensors was around $56.3 billion in 2010. In 2011, it was
around $62.8 billion. Global market for sensors is expected
to increase to $91.5 billion by 2016, at a compound annual
growth rate of 7.8%.
E. The Essential Component of IoT: Sensor Networks
We provide a brief introduction to sensor networks in this
section as it is the most essential component of the IoT. A
sensor network comprises one or more sensor nodes, which
417
communicate between themselves using wired and wireless

technologies. In sensor networks, sensors can be homogeneous
or heterogeneous. Multiple sensor networks can be connected
together through different technologies and protocols. One
such approach is through the Internet. The components and
the layered structure of a typical sensor network are discussed
in Section II-F.
We discuss how sensor networks and the IoT work together
in Section II-G. However, there are other technologies that
can complement the sensing and communication infrastructure
in IoT paradigm such as traditional ad-hoc networks. These
are clearly a different technology from sensor networks and
have many weaknesses. The differences are comprehensively
discussed in [39].
There are three main architectures in sensor networks: flat
architecture (data transfers from static sensor nodes to the sink
node using a multi-hop fashion), two-layer architecture (more
static and mobile sink nodes are deployed to collect data from
sensor nodes), and three-layer architecture (multiple sensor
networks are connected together over the Internet). Therefore,
IoT follows a three-layer architecture.
Most of the sensors deployed today are wireless. There
are several major wireless technologies used to build wireless
sensor networks: wireless personal area network (WPAN) (e.g.
Bluetooth), wireless local area network (WLAN) (e.g. Wi-Fi),
wireless metropolitan area network (WMAN) (e.g. WiMAX),
wireless wide area network (WWAN) (e.g. 2G and 3G networks), and satellite network (e.g. GPS). Sensor networks
also use two types of protocols for communication: non-IP
based (e.g: Zigbee and Sensor-Net) and IP-based protocols
(NanoStack, PhyNet, and IPv6).
The sensor network is not a concept that emerged with
the IoT. The concept of a sensor network and related research existed a long time before the IoT was introduced.
However, sensor networks were used in limited domains to
achieve specific purposes, such as environment monitoring
[40], agriculture [35], medical care [41], event detection [42],
structural health monitoring [43], etc. Further, there are three
categories of sensor networks that comprise the IoT [44]: body
sensor networks (BSN), object sensor networks (OSN), and
environment sensor networks (ESN).
Molla and Ahamed [8] identified ten challenges that need
to be considered when developing sensor network middleware solutions: abstraction support, data fusion, resource constraints, dynamic topology, application knowledge, programming paradigm, adaptability, scalability, security, and QoS
support. A comparison of different sensor network middleware
solutions is also provided based on the above parameters.
Several selected projects are also discussed in brief in order
to discover the approaches they take to address various challenges associated with sensor networks.
Some of the major sensor network middleware approaches
are IrisNet, JWebDust, Hourglass, HiFi, Cougar, Impala,
SINA, Mate, TinyDB, Smart Object, Agilla, TinyCubus,
TinyLime, EnviroTrack, Mires, Hood, and Smart Messages.
Some of the above approaches are surveyed in [8], [45].
A survey on web based wireless sensor architectures and
applications is presented in [46].
418

Fig. 3. Layered structure of a sensor network: These layers are identified based on the capabilities posed by the devices. In IoT, this layered architecture
may have additional number of sub layers as it is expected to comprises large verity of in sensing capabilities.
F. Layers in Sensor Networks

We have presented a typical structure of a sensor network
in Figure 3. It comprises the most common components in a
sensor network. As we have shown, with the orange coloured
arrows, data flows from right to left. Data is generated by
the low-end sensor nodes and high-end sensor nodes. Then,
data is collected by mobile and static sink nodes. The sink
nodes send the data to low-end computational devices. These
devices perform a certain amount of processing on the sensor
data. Then, the data is sent to high-end computational devices
to be processed further. Finally, data reaches the cloud where
it will be shared, stored, and processed significantly.
Based on the capabilities of the devices involved in a sensor
network, we have identified six layers. Information can be
processed in any layer. Capability means the processing, memory, communication, and energy capacity. Capabilities increase
from layer one to layer six. Based on our identification of
layers, it is evident that an ideal system should understand the
capability differences, and perform data management accordingly. It is all about efficiency and effectiveness. For example,
perform processing in the first few layers could reduce data
communication. However, devices in the first few layers do
not have a sufficient amount of energy and processing power
to do comprehensive data processing [47]. IoT research needs
to find more efficient and effective ways of data management,
such as collecting, modelling, reasoning, distributing.
G. Relationship Between Sensor Networks and IoT
In earlier sections we introduced both IoT and sensor
network concepts. In this section we explain the relationship between the two concepts. Previously, we argued that
sensor networks are the most essential components of the
IoT. Figure 4 illustrates the big picture. The IoT comprises
sensors and actuators. The data is collected using sensors.
Then, it is processed and decisions are made. Finally, actuators
perform the decided actions. This process is further discussed
in Section IV. Further, integration between wireless sensor
networks and the IoT are comprehensively discussed in [48].
The difference between sensor networks (SN) and the IoT is
largely unexplored and blurred. We can elaborate some of the
characteristics of both SN and IoT to identify the differences.
SN comprises of the sensor hardware (sensors and actuators), firmware and a thin layer of software. The IoT
comprises everything that SN comprises and further it comprises a thick layer of software such as middleware systems,
frameworks, APIs and many more software components. The
software layer is installed across computational devices (both
low and high-end) and the cloud.
From their origin, SNs were designed, developed, and used
for specific application purposes, for example, detecting bush
fire [44]. In the early days, sensor networks were largely
used for monitoring purposes and not for actuation [49]. In
contrast, IoT is not focused on specific applications. The
IoT can be explained as a general purpose sensor network
[50]. Therefore, the IoT should support many kinds of
applications. During the stage of deploying sensors, the IoT
would not be targeted to collect specific types of sensor
data, rather it would deploy sensors where they can be used
for various application domains. For example, company may
deploy sensors, such as pressure sensors, on a newly built
bridge to track its structural health. However, these sensors
may be reused and connect with many other sensors in
order to track traffic at a later stage. Therefore, middleware
solutions, frameworks, and APIs are designed to provide
generic services and functionalities such as intelligence,
semantic interoperability, context-awareness, etc. that are
required to perform communication between sensors and
actuators effectively.
Sensor networks can exist without the IoT. However, the IoT
cannot exist without SN, because SN provides the majority
of hardware (e.g. sensing and communicating) infrastructure
support, through providing access to sensors and actuators.
There are several other technologies that can provide access to sensor hardware, such as wireless ad-hoc networks.
However, they are not scalable and cannot accommodate
the needs of the IoT individually [39], though they can
complement the IoT infrastructure. As is clearly depicted
in Figure 4, SN are a part of the IoT. However, the IoT is
not a part of SN.
H. Characteristics of the IoT

In Section II-G, we highlighted the differences between sensor networks and the IoT. Further, we briefly explore the char-

Time considerations: The IoT could handle billions of

parallel and simultaneous events, due to the massive number
of interactions. Real-time data processing is essential.
Space considerations: The precise geographic location of a
object will be critical [54] as location plays a significant role
in context-aware computing. When the number of objects get
larger, tracking becomes a key requirement. Interactions are
highly dependent on their locations, their surroundings, and
presence of other entities (e.g. objects and people).
Everything-as-a-service: Due to the popularity of cloud
computing [55], consuming resources as a service [56]
such as Platform-as-a-Service (PaaS), Infrastructure-as-aService (IaaS), Software-as-a-Service (SaaS), has become
main stream. Everything-as-a-service [57] model is highly
efficient, scalable, and easy to use. IoT demands significant
amounts of infrastructure to be put in place in order to make
its vision a reality, where it would follow a community or
crowd based approach. Therefore, sharing would be essential,
where an everything-as-a-service model would suit mostly
sensing-as-a-service [5].

419

Fig. 4. Relationship between sensor networks and IoT.
acteristics of the IoT from a research perspective. Based on

previous research efforts we identified seven major characteristics in the IoT [4]: intelligence, architecture, complex system,
size considerations, time considerations, space considerations,
and everything-as-a-service. These characteristics need to be
considered when developing IoT solutions throughout all the
phases from design, development, implement and evaluation.
Intelligence: This means the application of knowledge.
First the knowledge needs to be generated by collecting
data and reasoning it. Transforming the collected raw data
into knowledge (high-level information) can be done by
collecting, modelling, and reasoning the context. Context can
be used to fuse sensor data together to infer new knowledge.
Once we have knowledge, it can be applied towards more
intelligent interaction and communication.
Architecture: IoT should be facilitated by a hybrid architecture which comprises many different architectures.
Primarily there would be two architectures: event driven [51]
and time driven. Some sensors produce data when an event
occurs (e.g. door sensor); the rest produce data continuously,
based on specified time frames (e.g. temperature sensor).
Mostly, the IoT and SN are event driven [52]. EventCondition-Action (ECA) rules are commonly used in such
systems.
Complex system: The IoT comprises a large number of
objects (sensors and actuators) that interact autonomously.
New objects will start communicating and existing ones will
disappear. Currently, there are millions of sensors deployed
around the world [53]. Interactions may differ significantly
depending on the objects capabilities. Some objects may
have very few capabilities, and as such store very limited
information and do no processing at all. In contrast, some
objects may have larger memory, processing, and reasoning
capabilities, which make them more intelligent.
Size considerations: It is predicted that there will be 50-100
billion devices connected to the Internet by 2020 [4]. The
IoT needs to facilitate the interaction among these objects.
The numbers will grow continuously and will never decrease.
Similar to the number of objects, number of interactions may
also increase significantly.
I. Middleware Support for IoT

As we mentioned at the beginning, the IoT needs to be
supported by middleware solutions. Middleware is a software
layer that stands between the networked operating system and
the application and provides well known reusable solutions
to frequently encountered problems like heterogeneity, interoperability, security, dependability [58]. The functionalities
required by IoT middleware solutions are explained in detail in
[4], [19], [20], [21], [29]. In addition, challenges in developing
middleware solutions for the IoT are discussed in [59]. We
present the summary of a survey conducted by Bandyopadhyay et al. [14]. They have selected the leading middleware
solutions and analyse them based on their functionalities,
each one offers, device management, interoperation, platform
portability, context-awareness, and security and privacy. Table
I shows the survey results. By the time we were preparing
this survey, some of the middleware solutions listed (i.e. GSN
and ASPIRE) were in the processing of extending towards
next generation solutions (i.e. EU FP7 project OpenIoT (20122014) [60]) by combining each others strengths.
J. Research Gaps
According to Table I, it can be seen that the majority of the
IoT middleware solutions do not provide context-awareness
functionality. In contrast, almost all the solutions are highly
focused on device management, which involves connecting
sensors to the IoT middleware. In the early days, contextawareness was strongly bound to pervasive and ubiquitous
computing. Even though there were some middleware solutions that provided an amount of context-aware functionality,
they did not satisfy the requirements that the IoT demands.
We discuss the issues and drawbacks with existing solutions,
in detail, in Section V. We discuss some of the research
directions in Section VII.
In this section, we introduced the IoT paradigm and highlighted the importance of context-awareness for the IoT. We
also learnt that context-awareness has not been addressed in
420
Middleware
DM
Hydra [61]
ISMB [62]
ASPIRE [63]
UBIWARE [64]
UBISOAP [65]
UBIROAD [66]
GSN [67]
SMEPP [68]
SOCRADES [69]
SIRENA [70]
WHEREX [71]
PP
Synonyms [81]: Circumstance, situation, phase, position,

posture, attitude, place, point; terms; regime; footing, standing, status, occasion, surroundings, environment, location,
dependence.
Definition by FOLDOC [82]: That which surrounds, and
gives meaning to, something else.
Definition by WordNet [83]: Discourse that surrounds a
language unit and helps to determine its interpretation
Definition by Longman [84]: The situation, events, or
information that are related to something and that help you
to understand it
TABLE I
I OT M IDDLEWARE C OMPARISON [14]
CA
SP
Legend: Device Management (DM), Interoperation (I), Platform

Portability (PP), Context Awareness (CA), Security & Privacy (SP)
existing IoT focused solutions, which motivates us to survey

the solutions in other paradigms to evaluate the applicability
of context-aware computing techniques toward IoT. In the next
section we discuss context-aware fundamentals that helps us
understand the in-depth discussions in the later sections.
III. C ONTEXT AWARENESS F UNDAMENTALS
This section discusses definitions of context and context
awareness, context-aware features, types of context and categorisation schemes, different levels and characteristics of
context-awareness, and finally, context management design
principles in the IoT paradigm.
A. Context-awareness Related Definitions
1) Definition of Context: The term context has been defined
by many researchers. Dey et al. [72] evaluated and highlighted
the weaknesses of these definitions. Dey claimed that the
definition provided by Schilit and Theimer [2] was based on
examples and cannot be used to identify new context. Further,
Dey claimed that definitions provided by Brown [73], Franklin
and Flachsbart [74], Rodden et al. [75], Hull et al. [76], and
Ward et al. [77] used synonyms to refer to context, such as
environment and situation. Therefore, these definitions also
cannot be used to identify new context. Abowd and Mynatt
[78] identified the five Ws (Who, What, Where, When, Why)
as the minimum information that is necessary to understand
context. Schilit et al. [79] and Pascoe [80] have also defined
the term context. Dey claimed that these definitions were too
specific and cannot be used to identify context in a broader
sense and provided a definition for context as follows:
Context is any information that can be used to characterise
the situation of an entity. An entity is a person, place, or
object that is considered relevant to the interaction between a
user and an application, including the user and applications
themselves [3].
We accept the definition of context provided by Abowd et al.
[3] to be used in this research work, because this definition can
be used to identify context from data in general. If we consider
a data element, by using this definition, we can easily identify
whether the data element is context or not. A number of
dictionaries have also defined and explained the word context:
In addition, Sanchez et al. [85] explained the distinction

between raw data and context information as follows:
Raw (sensor) data: Is unprocessed and retrieved directly
from the data source, such as sensors.
Context information: Is generated by processing raw sensor data. Further, it is checked for consistency and meta data
is added.
For example, the sensor readings produced by GPS sensors

can be considered as raw sensor data. Once we put the GPS
sensor readings in such a way that it represents a geographical
location, we call it context information. Therefore in general,
the raw data values produced by sensors can be considered as
data. If this data can be used to generate context information,
we identify these data as context. Therefore, mostly what we
capture from sensors are data not the context information.
Ahn and Kim [86] define context (also called compound
events) as a set of interrelated events with logical and timing
relations among them. They also define an event as an occurrence that triggers a condition in a target area. There are two
categories of events: discrete events and continuous events. If
the sampling rate is p:
Discrete events: An event that occurs at time t and t +
p, there are considered to have been two separate event
instances. (e.g. a door open, lights on, etc.)
Continuous events: An event instance lasting for at least
time p, where an event occurring at time t and t + p, cannot
be considered as two separate events. (e.g. raining, having a
shower, driving a car, etc.)
2) Definition of Context-awareness: The term context

awareness, also called sentient, was first introduced by Schilit
and Theimer [2] in 1994. Later, it was defined by Ryan et al.
[87]. In both cases, the focus was on computer applications
and systems. As stated by Abowd et al. [3], those definitions
are too specific and cannot be used to identify whether a given
system is a context-aware system or not. Therefore, Dey has
defined the term context-awareness as follows:
A system is context-aware if it uses context to provide relevant information and/or services to the user, where relevancy
depends on the users task. [3]
We accept the above definition on context-awareness to be
used in our research work, because we can use this definition
to identify context-aware systems from the rest. If we consider
a system, by using this definition we can easily identify
whether this system is a context-aware system or not. Context
awareness frameworks typically should support acquisition,
representation, delivery, and reaction [72]. In addition, there
are three main approaches that we can follow to build contextaware applications [88].
No application-level context model: Applications perform
all the actions, such as context acquisition, pre-processing,
storing, and reasoning within the application boundaries.
Implicit context model: Applications uses libraries, frameworks, and toolkits to perform context acquisition, preprocessing, storing, and reasoning tasks. It provides a standard design to follow that makes it easier to build the
applications quickly. However, still the context is hard bound
to the application.
Explicit context model: Applications uses a context management infrastructure or middleware solution. Therefore,
actions such as context acquisition, pre-processing, storing,
and reasoning lie outside the application boundaries. Context
management and application are clearly separated and can be
developed and extend independently.
3) Definition of Context Model and Context Attribute: We
adopt the following interpretations of context model and context attributes provided by Henricksen [89] based on Abowd
et al. [3] in our research work.
A context model identifies a concrete subset of the context
that is realistically attainable from sensors, applications and
users and able to be exploited in the execution of the task.
The context model that is employed by a given context-aware
application is usually explicitly specified by the application
developer, but may evolve over time [89].
A context attribute is an element of the context model
describing the context. A context attribute has an identifier,
a type and a value, and optionally a collection of properties
describing specific characteristics [89].
4) Definition of Quality of Context: There are number of
definitions and parameters that have been proposed in the
literature regarding quality of context (QoC). A survey on QoC
is presented in [16]. QoC is defined using a set of parameters
that expresses the quality of requirements and properties of the
context data. After evaluating a number of different parameter
proposals in the literature, [16] has defined QoC based on three
parameters: context data validity, context data precision, and
context data up-to-dateness. QoC are being used to resolve
context data conflicts. Further, they claim that QoC is depend
on quality of the physical sensor, quality of the context data,
and quality of the delivery process.
B. Context-aware Features
After analysing and comparing the two previous efforts conducted by Schilit et al. [79] and Pascoe [80], Abowd et al. [3]
identified three features that a context-aware application can
support: presentation, execution, and tagging. Even though,
the IoT vision was not known at the time these features are
identified, they are highly applicable to the IoT paradigm as
well. We elaborate these features from an IoT perspective.
Presentation: Context can be used to decide what information and services need to be presented to the user. Let
us consider a smart [22] environment scenario. When a user
enters a supermarket and takes their smart phone out, what
they want to see is their shopping list. Context-aware mobile
applications need to connect to kitchen appliances such as a
421
smart refrigerator [90] in the home to retrieve the shopping

list and present it to the user. This provides the idea of
presenting information based on context such as location,
time, etc. By definition, IoT promises to provide any service
anytime, anyplace, with anything and anyone, ideally using
any path/network.
Execution: Automatic execution of services is also a critical
feature in the IoT paradigm. Let us consider a smart home
[22] environment. When a user starts driving home from
their office, the IoT application employed in the house should
switch on the air condition system and switch on the coffee
machine to be ready to use by the time the user steps into
their house. These actions need to be taken automatically
based on the context. Machine-to-machine communication
is a significant part of the IoT.
Tagging: In the IoT paradigm, there will be a large number
of sensors attached to everyday objects. These objects will
produce large volumes of sensor data that has to be collected,
analysed, fused and interpreted [91]. Sensor data produced
by a single sensor will not provide the necessary information
that can be used to fully understand the situation. Therefore,
sensor data collected through multiple sensors needs to
be fused together. In order to accomplish the sensor data
fusion task, context needs to be collected. Context needs
to be tagged together with the sensor data to be processed
and understood later. Context annotation plays a significant
role in context-aware computing research. We also call this
tagging operation as annotation as well.
C. Context Types and Categorisation Schemes
Different researchers have identified context types differently based of different perspectives. Abowd et al. [3] introduced one of the leading mechanisms of defining context
types. They identified location, identity, time, and activity as
the primary context types. Further, they defined secondary
context as the context that can be found using primary context.
For example, given primary context such as a persons identity,
we can acquire many pieces of related information such as
phone numbers, addresses, email addresses, etc.
However, using this definition we are unable to identify the
type of a given context. Let us consider two GPS sensors
located in two different locations. We can retrieve GPS values
to identify the position of each sensor. However, we can
only find the distance between the two sensors by performing
calculations based on the raw values generated by the two
sensor. The question is, what is the category that distance
belongs to? is it primary or secondary? The distance is not
just a value that we sensed. We computed the distance by
fusing two pieces of context. The above definition does not
represent this accurately.
Thus, we define a context categorisation scheme (i.e. primary and secondary) that can be used to classify a given data
value (e.g. single data item such as current time) of context
in terms of an operational perspective (i.e. how the data was
acquired). However, the same data value can be considered
as primary context in one scenario and secondary context in
another. For example, if we collect the blood pressure level
of a patient directly from a sensor attached to the patient,
422

!

$"
$

%

$

"
!%

&

!

!

!

"

!

!

#
#

Fig. 5. Context categorisation in two different perspectives: conceptual and

operational. It shows why both operational and conceptual categorisation
schemes are important in IoT paradigm as the capture different perspectives.
it could be identified as primary context. However, if we

derive the same information from a patients health record
by connecting to the hospital database, we call it secondary
context. Therefore, the same information can be acquired
using different techniques. It is important to understand that
the quality, validity, accuracy, cost and effort of acquisition,
etc. may varied significantly based on the techniques used.
This would be more challenging in the IoT paradigm, because
there would be a large amount of data sources that can be used
to retrieve the same data value. To decide which source and
technique to use would be a difficult task. We will revisit this
challenge in Section VI. In addition, a similar type of context
information can be classified as both primary and secondary.
For example, location can be raw GPS data values or the
name of the location (e.g. city, road, restaurant). Therefore,
identifying a location as primary context without examining
how the data has been collected is fairly inaccurate. Figure 5
depicts how the context can be identified using our context
type definitions.
Primary context: Any information retrieved without using
existing context and without performing any kind of sensor
data fusion operations (e.g. GPS sensor readings as location
information).
Secondary context: Any information that can be computed
using primary context. The secondary context can be computed by using sensor data fusion operations or data retrieval
operations such as web service calls (e.g. identify the distance between two sensors by applying sensor data fusion
operations on two raw GPS sensor values). Further, retrieved
context such as phone numbers, addresses, email addresses,
birthdays, list of friends from a contact information provider
based on a personal identity as the primary context can also
be identified as secondary context.
We acknowledge location, identity, time, and activity as
important context information. The IoT paradigm needs to
consider more comprehensive categorisation schemes in a

hierarchical manner, such as major categories, sub categories
and so on. Operational categorisation schemes allow us to
understand the issues and challenges in data acquisition techniques, as well as quality and cost factors related to context.
In contrast, conceptual categorisation allows an understanding
of the conceptual relationships between context. We have to
integrate perspective in order to model context precisely. We
compare different context categorisation schemes in Table IV.
In addition to the two categorisation schemes we discussed
earlier there are several other schemes introduced by different
researchers focusing on different perspectives. Further, we
highlight relationships between different context categories
(also called context types) in different perspectives in Table II
and in Table III. These context categories are not completely
different from each other. Each category shares common
characteristics with the others. The similarities and difference
among categories are clearly presented in Table III. Further,
we have listed and briefly explained three major context categorisation schemes and their categories proposed by previous
researchers. In Table II, we present each categorisation effort
in chronological order from left to right.
Schilit et al. [79] (1994): They categorised context into three
categories using a conceptual categorisation based technique
on three common questions that can be used to determine
the context.
1) Where you are: This includes all location related information such as GPS coordinates, common names
(e.g. coffee shop, university, police), specific names
(e.g. Canberra city police), specific addresses, user
preferences (e.g. users favourite coffee shop).
2) Who you are with: The information about the people
present around the user.
3) What resources are nearby: This includes information
about resources available in the area where the user is
located, such as machinery, smart objects, and utilities.
Henricksen [89] (2003): Categorised context into four categories based on an operational categorisation technique.
1) Sensed: Sensor data directly sensed from the sensors,
such as temperature measured by a temperature sensor.
Values will be changed over time with a high frequency.
2) Static: Static information which will not change over
time, such as manufacturer of the sensor, capabilities
of the sensor, range of the sensor measurements.
3) Profiled: Information that changes over time with a low
frequency, such as once per month (e.g. location of
sensor, sensor ID).
4) Derived: The information computed using primary context such as distance of two sensors calculated using
two GPS sensors.
Van Bunningen et al. [95] (2005): Instead of categorising
context, they classified the context categorisation schemes
into two broader categories: operational and conceptual.
1) Operational categorisation: Categorise context based on
how they were acquired, modelled, and treated.
2) Conceptual categorisation: Categorise context based on
the meaning and conceptual relationships between the
context.
423
User
Computing (System)
Physical (Environment)
Historical
Social
Networking
Things
Sensor
Who (Identity)

Where (Location)

When (Time)
What (Activity)

Why
Sensed
Static
Profiled
Derived
Operational
Conceptual
Objective
Cognitive
External (Physical)
Internal (Logical)
Low-level (Observable)
High-level (Non-Observable)

(2011)
Yanwei [103]
(2011)
Liu [102]
(2010)
Rizou [101]
(2009)
Mei &
Easterbrook [100]
(2009)
Zhong [99]
(2007)
Chong [98]
(2007)
Guan [97]
(2006)
Miao and Yuan [96]
(2005)
Van Bunningen [95]
(2003)
Prekop &
Burnett [92],
Gustavsen [93],
Hofer [94]
(2003)
Henricksen [89]
(2000)
Chen and Kotz [6]
(1999)
Abowd [3]
(1997)
Ryan [87]
(1994)
Schilit [79]
Context Types
(1994)
Schilit [79]
TABLE II
D IFFERENT C ONTEXT C ATEGORISATION S CHEMES AND T HEIR S COPES

Based on the evaluation of context categorisation, it is evident that no single categorisation scheme can accommodate all
the demands in the IoT paradigm. We presented a comparison
between conceptual and operational categorisation schemes
in Table IV. To build an ideal context-aware middleware
solution for the IoT, different categorisation schemes need to
be combined together in order to complement their strengths
and mitigate their weaknesses.
D. Levels of Context Awareness and characteristics
Context awareness can be identified in three levels based
on the user interaction [104].
Personalisation: It allows the users to set their preferences,
likes, and expectation to the system manually. For example,
users may set the preferred temperature in a smart home
environment where the heating system of the home can
maintain the specified temperature across all rooms.
Passive context-awareness: The system constantly monitors the environment and offers the appropriate options to
the users so they can take actions. For example, when a user
enters a super market, the mobile phone alerts the user with
a list of discounted products to be considered.
Active context-awareness: The system continuously and
autonomously monitors the situation and acts autonomously.
For example, if the smoke detectors and temperature sensors

detect a fire in a room in a smart home environment, the

system will automatically notify the fire brigade as well as
the owner of the house via appropriate methods such as
phone calls.
In addition, Van Bunningen et al. [95] has identified comprehensively, and discussed, eight characteristics of context:
context 1) is sensed though sensors or sensor networks, 2) is
sensed by small and constrained devices, 3) originates from
distributed sources, 4) is continuously changing, 5) comes
from mobile objects 6) has a temporal character 7) has a spatial
character, 8) is imperfect and uncertain.
E. Context Awareness Management Design Principles
Martin et al. [105] have identified and comprehensively
discussed six design principles related to context-aware management frameworks (middleware). Further, Ramparany et al.
[106] and Bernardos et al. [107] have also identified several
design requirements. We summarise the findings below with
brief explanations. This list is not intended to be exhaustive.
Only the most important design aspects are considered.
Architecture layers and components: The functionalities need to be divided into layers and components in a
meaningful manner. Each component should perform a very
limited amount of the task and should be able to perform
independently up to a large extent.
424
2
2 3
2 2
2 2
2 2
2 2
(Q)
Internal (Logical)
External (Physical)
Cognitive
Objective
Profiled
Static
Sensed
2
3
3
2
2
2
2
(P)
Why
3
3 3
1 1 1
3 3 3 3
2 2 2 3 3
2 2 2 3 3 3
3 3 3 2 2 2 2
1 1 1 2 2 2 2
1 1 1 2 2 2 2
3 3 1 3 2 1 1
2 2 2 1 2 3 3
2 2 2 3 2 1 1
2 2 3 1 2 3 3
2 2 2 3 2 1 1
(Q) very high; 2 means
Conceptual
What (Activity)
When (Time)
Where (Location)
3
3
3
1
3
2
2
3
1
1
3
2
2
2
2
(P)
Operational
2
3 3
3 3
2 3
3 3
1 1
3 3
2 2
2 2
3 3
1 1
2 1
3 3
2 2
2 2
2 2
2 2
1 means
Scalability and extensibility: The component should be

able to added or removed dynamically. For example. new
functionalities (i.e. components) should be able to be add
without altering the existing components (e.g. Open Services
Gateway initiative). The component needs to be developed
according to standards across the solutions, which improves
scalability and extensibility (e.g. plug-in architectures).
Application programming interface (API): All the functionalities should be available to be accessed via a comprehensive easy to learn and easy to use API. This allows
the incorporation of different solutions very easily. Further,
API can be used to bind context management frameworks to
applications. Interoperability among different IoT solutions
heavily depends on API and their usability.
Debugging mechanisms and tools: Debugging is a critical task in any software development process. In the IoT
paradigm, debugging would be difficult due to the exponential number of possible alternative interactions. In order to
win the trust of the consumers, the IoT should prove its
trustworthiness. Integrated debug mechanisms inbuilt into
the framework will help to achieve this challenge. For
Who (Identity)
Sensor
Things
2
2 2
2 2
2 2
3 3
2 2
3 3
1 1
3 3
2 2
2 2
3 3
1 1
2 2
3 3
2 2
2 2
2 2
2 2
as (Q).
Derived
User
3
Computing (System)
3 3
3 2 2
Historical
3 2 2 2
Social
3 2 3 2 2
Networking
3 2 2 2 2
Things
3 2 1 2 2
Sensor
2 2 2 2 2
Who (Identity)
3 3 2 2 2
Where (Location)
3 3 3 2 3
When (Time)
3 2 2 2 2
What (Activity)
3 3 3 2 3
Why
1 1 1 2 1
Sensed
2 3 3 2 3
Static
2 2 2 2 2
Profiled
2 2 2 2 2
Derived
3 3 3 2 3
Operational
1 1 1 2 1
Conceptual
2 2 2 2 2
Objective
1 3 3 2 3
Cognitive
2 2 2 2 2
External (Physical)
2 2 2 2 2
Internal (Logical)
2 2 2 2 2
2 2 2 2 2
Notes: We denote row labels as (P) and column labels
(P) (Q) very low.
Networking
Social
Historical
Computing (System)
User
TABLE III
R ELATIONSHIP B ETWEEN D IFFERENT C ONTEXT C ATEGORIES
3
1 3
3 1 3
1
1 3
moderate; 3 means
example, the justifications behind the results produced by the

reasoners should be available to be evaluated to find possible
inaccuracies so further development can be carried out. Some
initial work in this area is presented in the Intelligibility
Toolkit [108].
Automatic context life cycle management: Context-aware
frameworks should be able to be understand by the available
context sources (i.e. physical and virtual sensors), their data
structure, and automatically built internal data models to
facilitate them. Further, raw context needs to be retrieved and
transformed into appropriate context representation models
correctly with minimum human intervention.
context model in-dependency: Context needs to be modelled and stored separately from context-aware framework
related code and data structures, which allows both parts to
be altered independently.
Extended, rich, and comprehensive modelling: Context
models should be able to extend easily. The IoT will need to
deal with enormous amount of devices, and will be required
to handle vast amounts of domain specific context. It also
needs to support complex relationships, constrains, etc. In
425
TABLE IV
C OMPARISON OF C ONTEXT C ATEGORISATION S CHEMES
Pros
Categorisation Schemes
Where, when, who,
what, objective
Cons
Provide a broader guide that helps to identify the

related context
Less comprehensive
Conceptual
User, computing,
physical, environmental,
time, social, networking,
things, sensors contexts
Why, cognitive
Sensed, static, profiled,

derived
Operational
Do not provide information about operational

aspects such as cost, time, complexity, techniques, and effort of data acquisition
Do not provide information about frequency of
update required
Allow to model mental reasoning behind context
Do not provide information about core context,

relationships between context or operational
Provide information about programming and coding

level
Provide information about context source and computational complexity
Allow to track information such as frequency of
update required, validation, quality, etc.
Provide information about cost and effort of data
acquisition
Weak in representing the relationship among

context
Difficult to classify context information due to
ambiguity. Same piece of data can belong to
different categories depending to the situation
(e.g. location can be derived as well as sensed)
Provide information about context sources and the

process of accessing data (e.g. whether more reasoning is required or not)
Provide information about cost and effort of data
acquisition
Provide information about computational complexity
Internal (physical),
internal (logical),
low-level (observable),
high-level
(non-observable)
More clear and structured method to organise context

More extensible and flexible
More comprehensive
Do not provide information about operational

Do not provide information about frequency of
update required
an ideal context-aware framework for the IoT, multiple different context representation models should be incorporated
together to improve their efficiency and effectiveness.
Multi-model reasoning: No single reasoning model can
accommodate the demands of the IoT. We will discuss
reasoning in Section IV-C. Each reasoning model has its
own strengths and weaknesses. An ideal framework should
incorporate multiple reasoning models together to complement each others strengths and mitigate their weaknesses.
Mobility support: In the IoT, most devices would be
mobile, where each one has a different set of hardware and
software capabilities. Therefore, context-aware frameworks
should be developed in multiple flavours (i.e. versions),
which can run on different hardware and software configurations (e.g. more capabilities for server level software and
less capabilities for mobile phones).
Share information (real-time and historic): In the IoT,
there is no single point of control. The architecture would
be distributed. Therefore, context sharing should happen at
different levels: framework-to-framework and frameworkto-application. Context model in-dependency has been discussed earlier and is crucial in sharing.
Resource optimisation: Due to the scale (e.g. 50 billion devices), a small improvement in data structures or processing
can make a huge impact in storage and energy consumption.
This stays true for any type of resource used in the IoT.
Monitoring and detect event: Events play a significant role
in the IoT, which is complement by monitoring. Detecting an
Weak in representing the relationship among

context
Difficult to classify context information due to
ambiguity. Same piece of data can belong to
different categories depending to the situation
(e.g. temperature can be physical or virtual
sensor)
event triggers an action autonomously in the IoT paradigm.

This is how the IoT will help humans carry out their dayto-day work easily and efficiently. Detecting events in real
time is a major challenge for context-aware frameworks in
the IoT paradigm.
IV. C ONTEXT L IFE C YCLE
A data life cycle shows how data moves from phase to
phase in software systems (e.g. application, middleware).
Specifically, it explains where the data is generated and where
the data is consumed. In this section we consider movement
of context in context-aware systems. Context-awareness is no
longer limited to desktop, web, or mobile applications. It
has already become a service: Context-as-a-Service (CXaaS)
[109]. In other terms, context management has become an
essential functionality in software systems. This trend will
grow in the IoT paradigm.
There are web-based context management services
(WCXMS) that provide context information management
throughout the contexts life cycle. Hynes et al. [109]
have classified data life cycles into two categories:
Enterprise Lifecycle Approaches (ELA) and Context
Lifecycle Approaches (CLA).
ELA are focused on context. However, these life cycles
are robust and well-established, based on industry standard
strategies for data management in general. In contrast, CLA
are specialised in context management. However, they are not
tested or standardised strategies as much as ELA. We have
426
selected ten popular data life cycles to analyse in this survey.

In the following list, 1-5 belong to ELA category and 6-10
belong to CLA category. Three dots (...) denotes reconnecting
to the first phase by completing the cycle. The right arrow
() denotes data transfer form one phase to another.
1) Information Lifecycle Management (ILM) [110]: creation and receipt distribution use maintenance
disposition ...
2) Enterprise Content Management (ECM) [111]: capture
manage store preserve deliver ...
3) Haydens Data Lifecycle [112]: collection relevance
classification handling and storage transmission and transportation manipulate, conversion and
alteration release backup retention destruction
...
4) Intelligence Cycle [113]: collection processing
analysis publication feedback ...
5) Boyd Control Loop (also called OODA loop) [114]:
observe orient decide act ...
6) Chantzara and Anagnostou Lifecycle [115]: sense (context provider) process (context broker) disseminate (context broker) use (service provider) ...
7) Ferscha et al. Lifecycle [116]: sensing transformation
representation rule base actuation ...
8) MOSQUITO [117]: context information discovery
context information acquisition context information
reasoning ...
9) WCXMS Lifecycle [109]: (context sensing context
transmission context acquisition ... ) context
classification context handling (context dissemination context usage context deletion context
request ... ) context maintenance context disposition ...
10) Baldauf et al. [10]: sensors raw data retrieval
reprocessing storage application.
In addition to the life cycles, Bernardos et al. [107] identified three phases in a typical context management system:
context acquisition, information processing, and reasoning and
decision. After reviewing the above life cycles, we derived an
appropriate (i.e. minimum number of phases but includes all
essential) context life cycle as depicted in Figure 6.
This context life cycle consists of four phases. First, context
needs to be acquired from various sources. The sources could
be physical sensors or virtual sensors (context acquisition).
Second, the collected data needs to be modelled and represent
according to a meaningful manner (context modelling). Third,
modelled data needs to be processed to derive high-level
context information from low-level raw sensor data (context
reasoning). Finally, both high-level and low-level context
needs to be distributed to the consumers who are interested
in context (context dissemination). The following discussion
is based on these four phases.
A. Context Acquisition
In this section we discuss five factors that need to be considered when developing context-aware middleware solutions in
the IoT paradigm. The techniques used to acquire context can
be varied based on responsibility, frequency, context source,
sensor type, and acquisition process.

Fig. 6. This is the simplest form of a context life cycle. These four steps are
essential in context management systems and middleware solutions. All the
other functions that may offer by systems are value added services.
1) Based on Responsibility: Context (e.g. sensor data)

acquisition can be primarily accomplished using two methods
[118]: push and pull. A comparison is presented in Table V.
Pull: The software component which is responsible for
acquiring sensor data from sensors make a request (e.g.
query) from the sensor hardware periodically (i.e. after
certain intervals) or instantly to acquire data.
Push: The physical or virtual sensor pushes data to the
software component which is responsible to acquiring sensor
data periodically or instantly. Periodical or instant pushing
can be employed to facilitate a publish and subscribe model.
2) Based on Frequency: Further, in the IoT paradigm,
context can be generated based on two different event types:
instant events and interval events VI.
Instant (also known as threshold violation): These events occur instantly. The events do not span across certain amounts
of time. Open a door, switch on a light, or animal enters
experimental crop field are some types of instant events. In
order to detect this type of event, sensor data needs to be
acquired when the event occurs. Both push and pull methods
can be employed.
Interval (also known as periodically): These events span a
certain period of time. Raining, animal eating a plant, or
winter are some interval events. In order to detect this type
of event, sensor data needs to be acquired periodically (e.g.
sense and send data to the software every 20 seconds). Both
push and pull methods can be employed.
3) Based on Source: In addition, context acquisition methods can be categorised into three categories [119] based on
where the context came from. A comparison is presented in
Table VII.
Acquire directly from sensor hardware: In this method, context is directly acquired from the sensor by communicating
with the sensor hardware and related APIs. Software drivers
and libraries need to be installed locally. This method is
typically used to retrieve data from sensors attached locally.
Most devices and sensors today require some amount of
driver support and can be connected via USB, COM, or serial
ports. However, wireless technologies are becoming popular
in the sensor community, which allows data transmission
427
TABLE V
C OMPARISON OF C ONTEXT A CQUISITION M ETHODS BASED ON R ESPONSIBILITY (P USH , P ULL )
Criteria
Pros
Push
Pull
Sensor hardware make the major decisions on sensing and

communication
Can be both instant or interval sensing and communication
Cons
Applicability
Decision on when to send data based on reasoning less

amount of data
Sensors are required to program when the requirements are
changed
Can be used when sensors know about when to send the data and
have enough processing power and knowledge to reason locally.
(e.g. event detection where one or small number of sensors can
reason and evaluate the conditions by their own without software
level complex data processing and reasoning.)
Software of the sensor data consumer makes the major

decisions on sensing and communication
Decision on when to collect data is based on reasoning
significant amount of data in software level
Can be both instant or interval sensing and communication
More communication bandwidth is required where software level has to send data requests to the sensors all the
time
Can be used when sensors do not have knowledge on when to

send the data to the consumer. (e.g. event detection where large
amount of data need to be collected, processed, and reasoned in
order to recognize the event.)
TABLE VI
C OMPARISON OF C ONTEXT A CQUISITION M ETHODS BASED ON F REQUENCY (I NSTANT, I NTERVAL )
Instant
Criteria
Pros
Interval
Save energy due to no redundant network communications

are involved
More accurate data can be gather as the network transmission would be triggered as soon as the conditions are met
Cons
Applicability
More knowledge is required to identify the conditions and

the satisfaction of the conditions
Hardware level (i.e. sensor) or software level should know
exactly what to look for
Difficult to detect events which require different types of
data from number of different sensors
Comparatively consume more energy for data processing
Can be used to detect frost events or heat events in agricultural

domain. In smart home domain, this method can be used to detect
some one entering to a room via door sensors. Ideally, applicable
for the situations where expected outcome is well-known by either
hardware level (i.e. sensors) or software level
without driver installations. In the IoT paradigm most objects

will communicate with each other via a wireless means.
Acquire through a middleware infrastructure: In this
method, sensor (context) data is acquired by middleware
solutions such as GSN. The applications can retrieve sensor
data from the middleware and not from the sensor hardware
directly. For example, some GSN instances will directly
access sensor hardware and rest of the GSN instances will
communicate with other GSN instances to retrieve data.
Acquire from context servers: In this method, context is
acquired from several other context storages (e.g. databases,
RSS (Really Simple Syndication) feeds, web services) via
different mechanisms such as web service calls. This mechanism is useful when the hosting device of the context-aware
application has limited computing resources. Resource-rich
context servers can be used to acquire and process context.
4) Based on Sensor Types: There are different types of
sensors that can be employed to acquire context. In general
Either sensors can be configured to sense and communicate

with data consumers in a predefined frequency or the sensor
data consumers can retrieve data explicitly from the sensors
in a predefined frequency
Sensors do not need to be intelligent/knowledge or have
significant processing and reasoning capabilities
Allows to understand the trends or behaviour by collecting
sensor data over time
May waste energy due to redundant data communication
Less accurate as the sensor readings can be change over the
interval between two data communications
Reasoning need to be done in software level by the data
consumer which will miss some occurrence of events due
to above inaccuracy
Can be used to collect data from temperature sensors for controlling air condition or measure air pollution where actions are
not event oriented but monitoring oriented. Ideally, applicable for
the situations where expected outcome is not known by either
hardware level (i.e. sensors) or software level
usage, the term sensor is used to refer to tangible sensor

hardware devices. However, among the technical community,
sensors are refer to as any data source that provides relevant
context. Therefore, sensors can be divided into three categories
[120]: physical, virtual, and logical. A comparison is presented
in Table VIII.
Physical sensors: These are the most commonly used type
of sensors and they are tangible. These sensors generate
sensor data by themselves. Most of the devices we use today
are equipped with a variety of sensor (e.g. temperature,
humidity, microphone, touch). A discussion on commonly
used sensor data types and sensors is presented in [121].
The data retrieved from physical sensors is called low-level
context. They are less meaningful, trivial, and vulnerable to
small changes. IoT solutions needs to understand the physical
world using imperfect, conflicting and imprecise data.
Virtual sensors: These sensors do not necessarily generate
sensor data by themselves. Virtual sensors retrieve data from
428
TABLE VII
C OMPARISON OF C ONTEXT A CQUISITION M ETHODS BASED ON S OURCE (D IRECT S ENSORS , M IDDLEWARE , C ONTEXT S ERVERS )
Criteria
Pros
Direct Sensor Access
Cons
Applicability
Through Middleware
Efficient as it allows direct communication with the sensors

Have more control over sensor configuration and data retrieval process
Significant technical knowledge is required including hardware level embedded device programming and configuring
Significant amount of time, effort,
cost involved
Updating is very difficult due to tight
bound between sensor hardware and
consumer application
Can be used for small scale scientific experiments. Can also be used for situation where
limited number of sensors are involved
Through Context Server
Easy to manage and retrieve context

as most of the management tasks are
facilitated by the middleware.
Can retrieve data faster with less effort and technical knowledge
Less resources required

Can retrieve data faster with less effort and technical knowledge
Require more resources (e.g. processing, memory, storage) as middleware

solutions need to be employed
Less control over sensor configuration
Moderately efficient as data need to
be retrieve through middleware
No control over sensor configuration

Less efficient as the context need to be
pulled from server over the network
IoT application will use this methods in

most cases. Can be used in situations where
large number of heterogeneous sensors are
involved
Can be used in situations where significant amount of context are required but
have only limited resources (i.e. cannot
employ context middleware solutions due
to resource limitations) that allows run the
consumer application
TABLE VIII
C OMPARISON OF C ONTEXT A CQUISITION M ETHODS BASED ON S ENSOR T YPES (P HYSICAL , V IRTUAL , L OGICAL )
Criteria
Pros
Physical Sensors
Cons
Applicability
Virtual Sensors
Error detection is possible and relatively easy

Missing value identification is also
relatively easy
Have access to low-level sensor configuration therefore can be more efficient
Hardware deployment and maintenance is costly

Have to deal with sensor and hardware level programming, design, development, test, debug
Provide less meaningful and low-level
raw sensor data
Can be used to collect physically observable

phenomenon such as light, temperature, humidity, gas, etc.
Logical Sensors
Provide moderately meaningful data

Provide high-level context information
Provided data are less processed
Do not need to deal with hardware
level tasks
Difficult to find errors in data

Filling missing values is not easy as
they are mostly non-numerical and
unpredictable
Can be used to collect information that cannot be measure physically such as calendar
details, email, chat, maps, contact details,
social networking related data, user preferences, user behaviour, etc.
many sources and publish it as sensor data (e.g. calendar,

contact number directory, twitter statuses, email and chat
applications). These sensors do not have a physical presence.
They commonly use web services technology to send and
receive data.
Logical sensors (also called software sensors): They combine physical sensors and virtual sensors in order to produce
more meaningful information. A web service dedicated to
providing weather information can be called a logical sensor.
Weather stations use thousands of physical sensors to collect
weather information. They also collect information from
virtual sensors such as maps, calendars, and historic data.
Finally, weather information is produced by combing both
physical and virtual sensors. In addition, the android mobile
operating system consists of a number of software sensors
Provide highly meaningful data

Provide high-level context information
Usually more accurate
Do not need to deal with hardware
level tasks
Difficult to find error in data
Filling missing values is not easy as
they are mostly non-numerical
Do not have control over data production process
License fees and other restrictions
may apply
Can be used to collect information that

are costly and impossible to collect directly through single physical sensor where
advance processing and fusing data from
multiple sensors are required (e.g. weather
information, activity recognition, location
recognition, etc.).
such as gravity, linear accelerometer, rotation vector, and

orientation sensors.
5) Based on Acquisition Process: There are three ways to
acquire context: sense, derive, and manually provided.
Sense: The data is sensed through sensors, including the
sensed data stored in databases (e.g. retrieve temperature
from a sensor, retrieve appointments details from a calendar).
Derive: The information is generated by performing computational operations on sensor data. These operations could be
as simple as web service calls or as complex as mathematical functions run over sensed data (e.g. calculate distance
between two sensors using GPS coordinates). The necessary
data should be available to apply any numerical or logical
reasoning technique.
Manually provided: Users provide context information manually via predefined settings options such as preferences (e.g.
understand that user doesnt like to receive event notifications
between 10pm to 6.00am). This method can be use to retrieve
any type of information.
B. Context Modelling
We discuss the basic definition of context modelling in
Section III-A3. Context modelling is also widely refereed to
as context representation. There are several popular context
modelling techniques [10], [122] used in context-aware computing. Before we present the discussion on context modelling techniques, lets briefly introduce context modelling
fundamentals. Context models can be static or dynamic. Static
models have a predefined set of context information that will
be collected and stored [103]. The requirements that need to be
taken into consideration when modelling context information
are identified and explained in [12] as heterogeneity and mobility, relationships and dependencies, timeliness (also called
freshness), imperfection, reasoning, usability of modelling
formalisms, and efficient context provisioning. Typically, there
are two steps in representing context according to a model:
Context modelling process: In the first step, new context
information needs to be defined in terms of attributes, characteristics, relationships with previously specified context,
quality-of context attributes and the queries for synchronous
context requests.
Organize context according to the model: In the second step,
the result of the context modelling step needs to be validated.
Then the new context information needs to be merged and
added to the existing context information repository. Finally,
the new context information is made available to be used
when required.
The first step performs the actual modelling of context.
However, the factors and parameters that are considered for
the modelling context are very subjective. It varies from one
solution to another. We use two examples to demonstrate the
variance. Currently, there is no standard to specify what type
of information needs to be considered in context modelling.
We discussed context categories proposed by the researcher
in Section III-C. Even though these categories provide highlevel guidelines towards choosing relevant context, choosing
specific context attributes is a subjective decision.
Example 1: MoCA [49] has used an object oriented approach to model context using XML. There are three sections
in the proposed context model: structural information (e.g.
attributes and dependencies among context types), behavioural
information (e.g. whether the context attribute has a constant
or variable value), and context-specific abstractions (e.g. contextual events and queries).
Example 2: W4 Diary [123] uses a W4 (who, what, where,
when) based context model to structure data in order to
extract high-level information from location data. For example, W4 represents context as tuples (e.g. Who: John,
What: walking:4km/h, Where: ANU, Canberra, When: 201301-05:9.30am).
In the IoT paradigm, context information has six states
[124]: ready, running, suspended, resumed, expired, and terminated. These states are also similar to the process states
429
in an operating system. They align context to an event. An

example scenario from the smart agriculture domain can be
used to explain the state transition of context.
Ready: Every context is in the ready state at the initial stage
(e.g. possible event can be an animal eating crop).
Suspended: When the context seems to be invalid temporally (e.g. sensors detect that animal stops eating crop
temporarily).
Resumed: When the context becomes valid from being
suspended (e.g. sensors detect animal starts to eat crop
again).
Expired: When the context has expired and further information is not available (e.g. sensor data has not been received
by the system for the last 60 seconds where all sensor data is
considered to be expired (based on policy) within 20 seconds
from the time it is collected).
Terminated: When the context is no longer valid (i.e.
inferred something else) and further information is not
available (e.g. sensors detects that animal moves away from
the crops).
The most popular context modelling techniques are surveyed in [6], [7]. These surveys discuss a number of systems
that have been developed based on the following techniques.
Each of the following techniques has its own strengths and
weaknesses. We discuss context modelling techniques at a
high-level. The actual implementations of these techniques
can vary widely depending on application domain (e.g. implementation details may differ from embedded environments to
mobile environments to cloud based environments). Therefore,
our focus is on conceptual perspective of each modelling
technique no on specific implementation. Our discussion is
based on the six most popular context modelling techniques:
key-value, markup schemes, graphical, object based, logic
based, and ontology based modelling. A comparison of these
models is presented in Table X.
1) Key-Value Modelling: It models context information as
key-value pairs in different formats such as text files and
binary files. This is the simplest form of context representation
among all the other techniques. They are easy to manage
when they have smaller amounts of data. However, key-value
modelling is not scalable and not suitable to store complex
data structures. Further, hierarchical structures or relationships
cannot be modelled using key-value pairs. Therefore, lack
of data structuring capability makes it difficult to retrieve
modelled information efficiently. Further, attaching meta information is not possible. The key-value technique is an
application oriented and application bounded technique that
suits the purpose of temporary storage such as less complex
application configurations and user preferences.
2) Markup Scheme Modelling (Tagged Encoding): It models data using tags. Therefore, context is stored within tags.
This technique is an improvement over the key-value modelling technique. The advantage of using markup tags is
that it allows efficient data retrieval. Further, validation is
supported through schema definitions. Sophisticated validation
tools are available for popular markup techniques such as
XML. Range checking is also possible up to some degree
for numerical values. Markup schemas such as XML are
widely used in almost all application domains to store data
430
TABLE IX
C OMPARISON OF S EMANTIC W EB O NTOLOGY L ANGUAGES (RDF(S), OWL(2))
RDF(S)
Cons
Pros
Provide basic elements to describe and organize

knowledge. Further, OWL is build on top of RDFS
Relatively simple
Faster processing and reasoning
Lack of inconsistency checking and reasoning

Limited expressiveness (e.g. no cardinality support)
OWL(2)
Improved version of RDFS. Therefore adaptability from RDF(S) to OWL is high

Increasing number of tools are supported
More expressive (e.g. larger vocabulary/constraints, rules, more meaningful)
Higher machine interoperability (e.g. strong syntax)
W3C approved standard for semantics (since 2004)
Comes in three versions (i.e. OWL light, OWL DL, OWL Full) where each one
has more expressive and reasoning power that previous
Relatively Complex
Low performance (e.g. require more computation power and time)
temporarily, transfer data among applications, and transfer

data among application components. In contrast, markup languages do not provide advanced expressive capabilities which
allow reasoning. Further, due to lack of design specifications,
context modelling, retrieval, interoperability, and re-usability
over different markup schemes can be difficult. A common
application of markup based modelling is modelling profiles.
Profiles are commonly developed using languages such as
XML. However, the concept of markup languages are not
restricted only to XML. Any language or mechanism (e.g.
JSON) that supports tag based storage allows markup scheme
modelling. An example of popular markup scheme modelling
is Composite Capabilities/Preference Profiles (CC/PP) [125].
There are a significant number of similar emerging applications such as ContextML [126] in context-aware computing.
Tuples are also used to model context [103].
3) Graphical Modelling: It models context with relationships. Some examples of this modelling technique are Unified
Modelling Language (UML) [127] and Object Role Modelling
(ORM) [128]. In terms of expressive richness, graphical
modelling is better than markup and key-value modelling as
it allows relationships to be captured into the context model.
Actual low-level representation of the graphical modelling
technique could be varied. For example, it could be a SQL
database, noSQL database, XML, etc. Many other extensions
have also been proposed and implemented using this technique
[89]. Further, as we are familiar with databases, graphical
modelling is a well known, easy to learn, and easy to use
technique. Databases can hold massive amounts of data and
provide simple data retrieval operations, which can be performed relatively quickly. In contrast, the number of different
implementations (i.e. different databases and other solutions)
makes it difficult with regards to interoperability. Further,
there are limitations on data retrieval mechanisms such as
SQL. In addition, sophisticated context retrieval requirements
may demand very complex SQL queries to be employed.
The queries can be difficult to create, use, and manage even
with the sophisticated tools that exist today. Adding context
information and changing the data structure is also difficult in
later stages. However, some of the recent trends and solutions
in the noSQL [129] movement allows these structure alteration issues to be overcome. Therefore, graphical modelling
techniques can be used as persistent storage of context.
4) Object Based Modelling: Object based (or object oriented) concepts are used to model data using class hierar-
chies and relationships. Object oriented paradigm promotes

encapsulation and re-usability. As most of the high-level
programming languages support object oriented concepts,
modelling can integrated into context-aware systems easily.
Therefore, object based modelling is suitable to be used as an
internal, non-shared, code based, run-time context modelling,
manipulation, and storage mechanism. However, it does not
provide inbuilt reasoning capabilities. Validation of object
oriented designs is also difficult due to the lack of standards
and specifications.
5) Logic Based Modelling: Facts, expressions, and rules
are used to represent information about the context. Rules are
used by other modelling techniques, such as ontologies, as
well. Rules are primarily used to express policies, constraints,
and preferences. It provides much more expressive richness
compared to the other models discussed previously. Therefore,
reasoning is possible up to a certain level. The specific structures and languages that can be used to model context using
rules are varied. However, lack of standardisation reduces the
re-usability and applicability. Furthermore, highly sophisticated and interactive graphical techniques can be employed
to develop logic based or rule based representations. As a
result, even non-technical users can add rules and logic to
the systems during run time. Logic based modelling allows
new high-level context information to be extracted using lowlevel context. Therefore, it has the capability to enhance other
context modelling techniques by acting as a supplement.
6) Ontology Based Modelling: The context is organised
into ontologies using semantic technologies. A number of
different standards (RDF, RDFS, OWL) and reasoning capabilities are available to be used depending on the requirement.
A wide range of development tools and reasoning engines
are also available. However, context retrieval can be computationally intensive and time consuming when the amount
of data is increased. According to many surveys, in contextaware computing and sensor data management, ontologies are
the preferred mechanism of managing and modelling context
despite its weaknesses. Due to its popularity and wider adaptation during the last five years in both academia and industry
we present a brief discussion on semantic modelling and
reasoning. However, our intention is not to survey semantic
technologies but to highlight the applicability of semantics
in a context-aware domain from an IoT perspective. Comprehensive and extensive amounts of information on semantic
technology are available in [130], [131], [132].
431
TABLE X
C OMPARISON OF C ONTEXT M ODELLING AND R EPRESENTATION T ECHNIQUES
Pros
Techniques
Cons
Key-Value
Simple
Flexible
Easy to manage when small in size
Markup
Scheme
Tagged
Encoding
(e.g. xml)
Graphical
(e.g.
databases)
Object
Based
Logic
Based
Ontology
Based
Flexible
More structured
Validation possible through schemas
Processing tools are available
Allows relationships modelling
Information retrieval is moderately
easier
Different standards and implementations are available.
Validation possible through constraints
Allows relationships modelling
Can be well integrated using programming languages
Allows to generate high-level context using low-level context
Simple to model and use
support logical reasoning
Support semantic reasoning
Allows more expressive representation of context
Strong validation
Application independent and allows
sharing
Strong support by standardisations
Fairly sophisticated tools available
Applicability
Strongly coupled with applications

Not scalable
No structure or schema
Hard to retrieve information
No way to represent relationships
No validation support
No standard processing tools are
available
Can be used to model limited amount of data

such as user preferences and application configurations. Mostly independent and non-related
pieces of information. This is also suitable for
limited data transferring and any other less complex temporary modelling requirements.
Application depended as there are

no standards for structures
Can be complex when many levels
of information are involved
Moderately difficult to retrieve information
Can be used as intermediate data organisation

format as well as mode of data transfer over
network. Can be used to decouple data structures
used by two components in a system. (e.g.
SensorML [133] for store sensor descriptions,
JSON as a format to data transfer over network)
Querying can be complex

Configuration may be required
Interoperability among different implementation is difficult
No standards but governed by design principles
Can be used for long term and large volume of

permanent data archival. Historic context can be
store in databases.
Hard to retrieve information

No standards but govern by design
principles
Lack of validation
No standards
Lack of validation
Strongly coupled with applications
Representation can be complex

Information retrieval can be complex and resource intensive
Khoo [134] has explained the evolution of the web in

four stages: basic Internet as Web 1.0, social media and
user generated content as web 2.0, semantic web as web 3.0
and IoT as web 4.0. In this identification, semantic web has
been given a separate phase to show its importance and the
significant changes that semantic technologies can bring to the
web in general.
Ontology is the main component in semantic technology
that allows it to model data. Based on the previous approaches
and survey [7], one of the most appropriate formats to manage
context is ontologies. Ontologies offer an expressive language
to represent the relationships and context. IT also provides
comprehensive reasoning mechanisms as well. Ontologies also
allow knowledge sharing and they decouple the knowledge
from the application and program codes [119].
There are several reasons to develop and use ontologies
in contrast to other modelling techniques. The most common
reasons are to [135], [136] share a common understanding of
the structure of information among people or software agents,
analyse domain knowledge, separate domain knowledge from
operational knowledge, enable reuse of domain knowledge,
Can be used to represent context in programming code level. Allows context runtime manipulation. Very short term, temporary, and mostly
stored in computer memory. Also support data
transfer over network.
Can be used to generate high-level context using
low-level context (i.e. generate new knowledge),
model events and actions (i.e. event detection),
and define constrains and restrictions.
Can be used to model domain knowledge and

structure context based on the relationships defined by the ontology. Rather than storing data
on ontologies, data can be stored in appropriate
data sources (i.e. databases) while structure is
provided by ontologies.
high-level knowledge inferring, and make domain assumptions

explicit. Due to the dynamic nature, the IoT middleware solutions should support applications which are not even known at
the middleware design-time. Ontologies allow the integration
of knowledge on different domains into applications when
necessary.
Studer et al. [137] defined the concept of ontology as
follows. An ontology is a formal, explicit specification of
a shared conceptualisation. A conceptualisation refers to an
abstract model of some phenomenon in the world by having
identified the relevant concepts of that phenomenon. Explicit
means that the type of concepts used, and the constraints
on their use are explicitly defined. For example, in medical
domains, the concepts are diseases and symptoms, the relations between them are causal and a constraint is that a
disease cannot cause itself. Formal refers to the fact that
the ontology should be machine readable, which excludes
natural language. Shared reflects the notion that an ontology
captures consensual knowledge, that is, it is not private to
some individual, but accepted by a group. Another acceptable
definition has been presented by Noy and McGuinness [136].
432
Further ontologies are discussed extensively as principles,

methods, and applications in perspective [138].
Some of the requirements and objectives behind designing an ontology are simplicity, flexibility and extensibility,
generality, and expressiveness [139]. In addition, some of
the general requirements in context modelling and representation are unique identification, validation, reuse, handling
uncertainty, and incomplete information [11]. A further eight
principles for developing ontologies are identified by Korpipaa
and Mantyjarvi [140] as: domain, simplicity, practical access,
flexibility and expandability, facilitate inference, genericity,
efficiency, and expressiveness.
Ontologies consists of several common key components
[141], [142] such as individuals, classes, attributes, relations,
function terms, restrictions, rules, axioms, and events. Furthermore, there are two steps in developing ontologies. First, the
domain and scope need to be clearly defined. Then existing
ontologies need to be reviewed to find the possibilities of
leverage existing in ontologies. One of the main goals of
ontologies is the reusability of shared knowledge. By the time
this survey was prepared, there were several popular domains
that design, develop, and use ontologies. Sensor domain is one
of them. A survey of the semantic specification of sensors
is presented in [143]. They have evaluated and compared a
number of ontologies and their capabilities.
There are several popular semantic web ontology languages
that can be used to develop ontologies: RDF [144], RDFS
[145], OWL [146]. The current recommendation is OWL 2
which is an extended version of OWL. A significant amount
of OWL usage has been noticed in the context modelling ad
reasoning domain [11]. It further emphasises the requirement
of having the modelling language, reasoning engines, and
mechanism to define rules as a bundle, rather than choosing
different available options arbitrarily, to get the real power of
semantic technologies. SWRL is one of the available solutions
to add rules in OWL [12]. SWRL is not a hybrid approach
as it is fully integrated into ontological reasoning. In contrast,
when the amount of data becomes larger and structure becomes complex, ontologies can becomes exceedingly complex
causing the reasoning process to be resource intensive and
slow. However, some of the main reasons to choose OWL as
the context modelling mechanism are [119], [142].
W3C strongly supports the standardisation of OWL. Therefore, a variety of development tools are available for integrating and managing OWL ontologies, which makes it easier
to develop and share.
OWL allows interoperability among other context-aware
systems. These features, such as classes, properties and
constraints, and individuals are important for supporting
ontology reuse, mapping and interoperability.
OWL supports a high-level of inference / reasoning support.
OWL is more expressive. For example, it provides cardinality constraints, which enables imposing additional restrictions on the classes.
We compare the two most popular web ontology languages,
RDF(S) and OWL(2) in Table IX, to highlight the fundamental
differences.
After evaluating several context modelling techniques, it
was revealed that incorporating multiple modelling techniques
is the best way to produce efficient and effective results, which

will mitigate each others weaknesses. Therefore, no single
modelling technique is ideal to be used in a standalone fashion.
There is a strong relationship between context modelling and
reasoning. For example, some reasoning techniques prefer
some modelling techniques. However, it should not limit the
employability of different context reasoning and modelling
techniques together. In the next section we discuss reasoning
context-aware computing.
C. Context Reasoning Decision Models

Context reasoning can be defined as a method of deducing new knowledge, and understanding better, based on the
available context [147]. It can also be explained as a process
of giving high-level context deductions from a set of contexts
[97]. The requirement of reasoning also emerged due to two
characteristics of raw context: imperfection (i.e. unknown, ambiguous, imprecise, or erroneous) and uncertainty. Reasoning
performance can be measured using efficiency, soundness,
completeness, and interoperability [11]. Reasoning is also
called inferencing. Contest reasoning comprises several steps.
Broadly we can divide them into three phases [148].
Context pre-processing: This phase cleans the collected
sensor data. Due to inefficiencies in sensor hardware and
network communication, collected data may be not accurate
or missing. Therefore, data needs to be cleaned by filling
missing values, removing outliers, validating context via
multiple sources, and many more. These tasks have been
extensively researched by database, data mining, and sensor
network research communities over many years.
Sensor data fusion: It is a method of combining sensor
data from multiple sensors to produce more accurate, more
complete, and more dependable information that could not
be achieve through a single sensor [149]. In the IoT, fusion is
extremely important, because there will be billions of sensors
available. As a result, a large number of alternative sources
will exist to provide the same information.
Context inference: Generation of high-level context information using lower-level context. The inferencing can be
done in a single interaction or in multiple interactions.
Revisiting an example from a different perspective, W4 Diary
[123] represented context as tuples (e.g. Who: John, What:
walking:4km/h, Where: ANU,Canberra, When: 2013-0105:9.30am). This low-level context can be inferred through
a number of reasoning mechanisms to generate the final results. For example, in the first iteration, longitude and latitude
values of a GPS sensor may be inferred as PurplePickle
cafe in canberra. In the next iteration PurplePickle cafe in
canberra may be inferred as Johns favourite cafe. Each
iteration gives more accurate and meaningful information.
There are a large number of different context reasoning

decision models, such as decision tree, naive Bayes, hidden
Markov models, support vector machines, k-nearest neighbour,
artificial neural networks, Dempster-Shafer, ontology-based,
rule-based, fuzzy reasoning and many more. Most of the
models originated and are employed in the fields of artificial
intelligence and machine learning. Therefore, these models are

433

$#

$"

!
"

!
Fig. 7. (a) Counts of model types used in 109 of 114 reviewed context-aware applications. (b) Counts for 50 recognition applications; classifiers are used
most often for applications that do recognition [108].
not specific to context-reasoning but commonly used across

many different fields in computing and engineering.
We present the results of a survey conducted by Lim and
Dey [108] in Figure 7. They have investigated the popularity
of context reasoning decision models. The survey is based
on literature from three major conferences over five years:
Computer-Human Interaction (CHI) 2003-2009, Ubiquitous
Computing (Ubicomp) 2004-2009, and Pervasive 2004-2009.
In the IoT paradigm, there are many sensors that sense
and produce context information. The amount of information
that will be collected by over 50 billion sensors is enormous.
Therefore, using all this context for reasoning in not feasible
for many reasons, such as processing time, power, storage,
etc. Furthermore, Guan et al. [97] has proved that using
more context will not necessarily improve the accuracy of
the inference in a considerable manner. They have used two
reasoning models in their research: back-propagation neural
networks and k-nearest neighbours. According to the results,
93% accuracy has been achieved by using ten raw context.
Adding 30 more raw context to the reasoning model has
increased the accuracy by only 1.63%. Therefore, selecting
the appropriate raw context for reasoning is critical to infer
high-level context with high accuracy.
Context reasoning has been researched over many years.
The most popular context reasoning techniques (also called
decision models) are surveyed in [11], [12], [147]. Our intention in this paper is not to survey context reasoning techniques
but to briefly introduce them so it will help to understand and
appreciate the role of context reasoning in the IoT paradigm.
We classify context reasoning techniques broadly into six
categories: supervised learning, unsupervised learning, rules,
fuzzy logic, ontological reasoning and probabilistic reasoning.
A comparison of these techniques is presented in Table XI
1) Supervised learning: In this category of techniques, we
first collect training examples. Then we label them according
to the results we expect. Then we derive a function that can
generate the expected results using the training data. This
technique is widely used in mobile phone sensing [150] and
activity recognition [151]. Decision tree is a supervised learning technique where it builds a tree from a dataset that can be
used to classify data. This technique has been used to develop
a student assessment system in [152]. Bayesian Networks is
a technique based on probabilistic reasoning concepts. It uses
directed acyclic graphs to represent events and relationships
among them. It is a widely used technique in statistical

reasoning. Example applications are presented in [141], [153].
Bayesian networks are commonly used in combining uncertain
information from a large number of sources and deducing
higher-level contexts. Artificial neural networks is a technique
that attempts to mimic the biological neuron system. They are
typically used to model complex relationships between inputs
and outputs or to find patterns in data. Body sensor networks
domain has employed this technique for pervasive healthcare
monitoring in [154]. Support vector machines are widely used
for pattern recognition in context-aware computing. It has
been used to detect activity recognition of patients in the
healthcare domain [155] and to learn situations in a smart
home environment [156].
2) Unsupervised learning: This category of techniques
can find hidden structures in unlabelled data. Due to the
use of no training data, there is no error or reward signal
to evaluate a potential solution. Clustering techniques such
as K-Nearest Neighbour is popularly used in context-aware
reasoning. Specifically, clustering is used in low-level (sensor
hardware level) sensor network operations such as routing
and high level tasks such as indoor and outdoor positioning
and location [157]. Unsupervised neural network techniques
such as Kohonen Self-Organizing Map (KSOM) are used to
classify incoming sensor data in a real-time fashion [158].
Noise detection and outlier detection are other applications
in context-aware computing. Applications of unsupervised
learning techniques in relation to body sensor networks are
surveyed in [154]. The unsupervised clustering method has
been employed to capturing user contexts by dynamic profiling
in [159].
3) Rules: This is the simplest and most straightforward
methods of reasoning out of all of them. Rules are usually
structure in an IF-THEN-ELSE format. This is the most
popular method of reasoning according to Figure 7. It allows
the generation of high level context information using low
level context. Recently, rules have been heavily used when
combined with ontological reasoning [160], [161], [162].
MiRE [163] is a minimal rule engine for context-aware mobile
devices. Most of the user preferences are encoded using rules.
Rules are also used in event detection [164], [165]. Rules are
expected to play a significant role in the IoT, where they are
the easiest and simplest way to model human thinking and
reasoning in machines. PRIAMOS [166] has used semantic
434
TABLE XI
C OMPARISON OF C ONTEXT R EASONING D ECISION M ODELLING T ECHNIQUES
Pros
Techniques
Cons
Supervised
Learning
(Artificial neural
network,
Bayesian
Networks,
Case-based
reasoning,
Decision tree
learning, Support
vector machines)
Unsupervised
Learning
(Clustering,
k-Nearest
Neighbour)
Rules
Probabilistic logic
(Dempster-Shafer,
hidden Markov
Models, naive
Bayes)
Ontology based
(First-Order
Predicate Logic)
Fuzzy Logic
Fairly accurate
Number of alternative models are
available
Have mathematical and statistical
foundation
No training data required

No need to know the possible outcome
Simple to define
Easy to extend
Less resource (e.g. processing, storage) intensive
Allow more natural representation
Simple to define
Easy to extend
Less resource (e.g. processing, storage) intensive
Can handle uncertainty
Allow complex reasoning
Allow complex representation
More meaningful results
Validation and quality checking is
possible
Can reason both numerical and textual data
Allows to combine evidence
Can handle unseen situations
Alternative models are available
Can handle uncertainty
provide moderately meaningful results
Applicability
Require significant amount of

data
Every data element need to be
converted in to numerical values
Selecting feature set could be
challenging
Can be more resource intensive
(processing, storage, time)
less semantic so less meaningful
Training data required
Models can be complex
Difficult to capture existing
knowledge
For situation where the feature set is easily

identifiable, possible out comes are known,
and large data sets (for training as well) are
available in numerical terms. (For example:
activity recognition, missing value identification)
Models can be complex

Less semantic so less meaningful
Difficult to validate
Outcome is not predictable
Can be more resource intensive
(processing, storage, time)
For situations where possible out comes are

not known (For example: unusual behaviour
detection, analysing agricultural fields to
identify appropriate location to plant a specific type of crop)
Should define manually

Can be error prone due to manual
work
No validation or quality checking
For situations where raw data elements need

to be converted in to high level context
information. Suitable to be used to define
events.
Should define manually

Can be error prone due to manual
work
No validation or quality checking
May reduce the quality (e.g. precision) of the results due to natural representation
For situation where low-level context need

to be converted in to high-level more natural
context information. This type of simplification will make it easy to process further.
For example, control automated irrigation
system where water will be released when
the system detect the soil is dry
Data need to be modelled in

a compatible format (e.g. OWL,
RDF)
Limited numerical reasoning
Low performance (e.g. require
more computation power and
time)
For situations where knowledge is critical. For example, store and reason domain
knowledge about agricultural domain. It allows the context information to be store
according to the ontology structure and automatically reason later when required
Should know the probabilities

Reason numerical values only
For situations where probabilities are

known and combing evidence from different
sources are essential. For example, evidence
produced from a camera, infra-red sensors,
acoustics sensor, and motion detector can be
combined to detect a wind animal infiltrate
to a agricultural field
rules to annotate sensor data with context information. Application of rule based reasoning is clearly explained in relation
to context-aware I/O control in [167].
4) Fuzzy logic: This allows approximate reasoning instead
of fixed and crisp reasoning. Fuzzy logic is similar to probabilistic reasoning but confidence values represent degrees of
membership rather than probability [168]. In traditional logic
theory, acceptable truth values are 0 or 1. In fuzzy logic partial
truth values are acceptable. It allows real world scenarios to
be represented more naturally; as most real world facts are
not crisp. It further allows the use of natural language (e.g.
temperature: slightly warm, fairly cold) definitions rather than
exact numerical values (e.g. temperature: 10 degrees Celsius).
In other words it allows imprecise notions such as tall, short,
dark, trustworthy and confidence to be captured, which is
critical in context information processing. In most cases, fuzzy
reasoning cannot be used as a standalone reasoning technique.
It is usually used to complement another techniques such as

rules based, probabilistic or ontological reasoning. Gaia [169]
has used fuzzy logic in context providers to handle uncertainty.
Several examples of applying fuzzy logic to represent context
information are presented in [170], [171].
5) Ontology based: It is based on description logic, which
is a family of logic based knowledge representations of
formalisms. Ontological reasoning is mainly supported by
two common representations of semantic web languages:
RDF(S) [144] and OWL(2) [146]. We discussed ontology
based modelling in Section IV-B6. Semantic web languages
are also complemented by several semantic query languages:
RDQL, RQL, TRIPLE and number of reasoning engines:
FACT [172], RACER, Pellet [173]. Rules such as SWRL
[160] are increasingly popular in ontological reasoning. The
advantage of ontological reasoning is that it integrates well
with ontology modelling. In contrast, a disadvantage is that
ontological reasoning is not capable of finding missing values

or ambiguous information where statistical reasoning techniques are good at that. Rules can be used to minimise
this weakness by generating new context information based
on low-level context. Missing values can also be tackled
by having rules that enable missing values to be replaced
with suitable predefined values. However, these mechanism
will not perform accurately in highly dynamic and uncertain
domains. Ontological reasoning is heavily used in a wide
range of applications, such as activity recognition [151],
hybrid reasoning [151], and event detection [165]. A survey
on semantic based reasoning is presented in [147]. It also
compares a number of context aware frameworks based on
modelling technique, reasoning techniques, and architectures
used in their systems. Comprehensive and extensive amounts
of information on semantic technology are available in [130],
[131], [132]. In addition, a semantic based architecture for
sensor data fusion is presented in [174], [175], [176].
6) Probabilistic logic: This category of techniques allows
decisions to be made based on probabilities attached to the
facts related to the problem. It can be used to combine
sensor data from two different sources. Further, it can be
used to identify resolutions to conflicts among context. Most
often these techniques are used to understand occurrence of
events. Probabilistic logic has been used in [168] to encode
access control policies. Dempster-Shafer, which is based on
probabilistic logic, allows different evidence to be combined
to calculate the probability of an event. Dempster-Shafer is
commonly used in sensor data fusion for activity recognition.
In [171], [177], it has been used to understand whether
there is a meeting in the room. Other example applications
are presented in [178], [179]. hidden Markov Models [180]
are also a probabilistic technique that allows state to be
represented using observable evidence without directly reading
the state. For example, it provides a method to bridge the
gap between raw GPS sensor measurements and high level
information such as a user destination, mode of transportation,
calendar based observable evidence such as user calendar,
weather, etc. hidden Markov Models are commonly used in
activity recognition in context-aware domains. For example,
it has been used to learn situation models in a smart home
[156].
Up to now, we have presented and discussed a number of
context modelling and reasoning techniques. However, it is
clear that each technique has its own strengths and weakness.
No single technique can be used to accomplish perfect results.
Therefore, the best method to tackle the problem of context
awareness it to combine multiple models in such a way that,
as a whole, they reduce weaknesses by complementing each
other. For example, Alternative Context Construction Trees
(ACCT) [181] is an approach that enables the concurrent
evaluation and consolidation of different reasoning models
such as logic rules, Bayesian networks and CoCoGraphs [182].
There are two reasons that context information can become
uncertain, as discussed in V-A16. Therefore, employing or
incorporating strategies that can reason under uncertainty such
as Bayesian networks, Dempster-Shafer or fuzzy logic is
essential in such situations. The process of how the multiple
techniques can be combined together is presented in [12],
435
[183]. We briefly explain the hybrid context modelling and

reasoning approach as follows.
At the lowest level, statistical techniques can be used to fuse
sensor data. Then, fuzzy logic can be employed to convert
fixed data in to more natural terms. In the future, DempsterShafer can be used to combine sensor data from different
sources. In addition, machine learning techniques, such as
support vector machines and artificial neural networks, can
be used for further reasoning. After completing statistical
reasoning, the high level data can be modelled using semantic
technologies such as ontologies. Ontological reasoning can be
applied to infer additional context information using domain
knowledge at the higher level. A similar process is explained
in detail in [183].
D. Context Distribution
Context distribution is a fairly straightforward task. It
provides methods to deliver context to the consumers. From
the consumer perspective this task can be called context
acquisition, where the discussion we presented in Section
IV-A is completely applicable. Therefore all the factors we
discussed under context acquisition need to be considered for
context distribution as well. Other than that there are two other
methods to that are used commonly in context distribution:
Query: Context consumer makes a request in terms of a
query, so the context management system can use that query
to produce results.
Subscription (also called publish / subscribe): Context consumer can be allowed to subscribe with a context management system by describing the requirements. The system
will then return the results periodically or when an event
occurs (threshold violation). In other terms, consumers can
subscribe for a specific sensor or to an event. However, in
underline implementations, queries may also use to define
subscriptions. Further, this method is typically use in real
time processing.
V. E XISTING R ESEARCH P ROTOTYPES AND S YSTEMS
In this section, first we present our evaluation framework
and then we briefly discuss some of the most significant
projects and highlight their significance. Later, we identify
the lessons we can learn from them towards context-aware
development in the IoT paradigm in Section VI. The projects
are discussed in the same order as in Table XIII. Our taxonomy
is summarized in Table XII.
A. Evaluation Framework
We used abbreviations as much as possible to make sure
that the structure allowed all 50 projects to be presented in a
single page, which enables the readers to analyse and identify
positive and negative patterns that we have not explicitly
discussed. In Table XIII, we use a dash () symbol across
all columns to denote that the functionality is either missing
or not mentioned in related publications that are available.
In order to increase the readability, we have numbered the
columns of the Table XIII corresponding to the taxonomy
numbered below. Our taxonomy and several other features
that will provide additional value in IoT solutions are visually
illustrated in Figure 8.
436
1) Project Name: This is the name given to the project

by the authors of the related publications. Most of the project
names are abbreviations that are used to refer to the project.
However, some project do not have an explicit project name,
here we used a dash () symbol.
2) Citation: We provide only one citation due to space
limitations. Other citations are listed under each projects
descriptions and highlights in Section V.
3) Year: Table XIII is ordered according to chronological
order (i.e. from oldest to newest) based on the year of
publication.
4) Project Focus: Based on our evaluation, each project
has its own focus on whether to build a system, a toolkit, or a
middleware solution. The following abbreviations are used to
denote the focus: system (S), toolkit (T), and middleware (M).
Systems focus on developing an end-to-end solution where
it involves hardware, software and application layer. Systems
cannot be used as middleware. It is designed to provide one
or a few tasks. Building different functionalities on top of the
system is not an option. Systems are designed and developed
for a use by the end users. Toolkits are not designed to be used
by the end users. They are employed by system, application,
and middleware developers. They provide very specific functionalities. Toolkits are usually designed according to wellknown design principles and standards and always released
with proper documentation that shows how to use them at
programming code level. Middleware [58] can be explained as
a software layer that lies between the hardware and application
layers. It provides reusable functionalities that are required
by the application to meet complex customer requirements.
They are usually built to address common issues in application
development such as heterogeneity, interoperability, security,
and dependability. A goal of middleware is to provide a set
of programming abstractions to help software development
where heterogeneous components need to be connected and
communicate together. Middleware is designed to be used by
application developers, where the middleware solution handles
most of the common functionalities leaving more time and
effort for the application developers to deal with application
functionalities.
5) Modelling: This has been discussed in detail in Section
IV-B. We use the following abbreviations to denote the context
modelling techniques employed by the project: key-value
modelling (K), markup Schemes (M), graphical modelling (G),
object oriented modelling (Ob), logic-based modelling (L),
and ontology-based modelling (On).
6) Reasoning: This has been discussed in detail in Section
IV-C. We use the following abbreviations to denote the context
reasoning techniques employed by the project: supervised
learning (S), un-supervised learning (U), rules (R), fuzzy logic
(F), ontology-based (O), and probabilistic reasoning (P). The
symbol () is used where reasoning functionality is provided
but the specific technique is not mentioned.
7) Distribution: This has been discussed in detail in Section IV-D. We use the following abbreviations to denote
the context distribution techniques employed by the project:
publish/subscribe (P) and query (Q).
8) Architecture: This varied widely from one solution to
another. Architecture can be classified into different categories
based on different perspectives. Therefore, there is no common

classification scheme that can be used for all situations. We
consider the most significant architectural characteristics to
classify the solution. Different architectural styles are numbered as follows. (1) Component based architecture where
the entire solution is based on loosely coupled major components, which interact each other. For example, Context
Toolkit [72] has three major components which perform the
most critical functionalities of the system. (2) Distributed
architecture enables peer-to-peer interaction in a distributed
fashion, such as in Solar [184]. (3) Service based architecture
where the entire solution consists of several services working
together. However, individual access to each service may not
be provided in solutions such as Gaia [168]. (4) Node based
architecture allows to deployment of pieces of software with
similar or different capabilities, which communicate and collectively process data in sensor networks [85]. (5) Centralised
architecture which acts as a complete stack (e.g. middleware)
and provides applications to be developed on top of that, but
provides no communication between different instances of the
solution. (6) Client-server architecture separates sensing and
processing from each other, such as in CaSP [185].
9) History and Storage: Storing context history is critical
[186] in both traditional context-aware computing and the IoT.
Historic data allows sensor data to be better understood. Even
though most of the IoT solutions and applications are focused
on real time interaction, historic data has its own role to play.
Specifically, it allows user behaviours, preferences, patterns,
trends, needs, and many more to be understood. In contrast,
due to the scale of the IoT, storing all the context for the long
term may not feasible. However, storage devices are getting
more and more powerful and cheap. Therefore, it would be
a tradeoff between cost and understanding. The symbol ()
is used denote that context history functionality is facilitated
and employed by the project.
10) Knowledge Management: This functionality is broader
than any others. Most of the tasks that are performed by
IoT middleware solutions require knowledge in different perspectives, such as knowledge on sensors, domains, users,
activities, and many more. One of the most popular techniques
to represent knowledge in context-aware computing is using
ontologies. However, several other techniques are also available such as rules. Knowledge can be used for tasks such as
automated configuration of sensors to IoT middleware, automatic sensor data annotation, reasoning, and event detection.
The symbol () is used to denote that knowledge management
functionality is facilitated and employed by the project in some
perspective.
11) Event Detection: This is one of the most important
functionalities in IoT solutions. IoT envisions machine-tomachine (M2M) and machine-to-person communication. Most
of these interactions are likely to occur based on an event.
Events can referred to many things, such as an observable
occurrence, phenomenon, or an extraordinary occurrence. We
define one or more conditions and identify it as an occurrence
of an event once all the defined conditions are satisfied. In
the IoT, sensors collect data and compare it with conditions to
decide whether the data satisfies the conditions. An occurrence
event is also called a event trigger. Once an event has
been triggered, a notification or action may be executed. For

example, detecting current activity of a person or detecting a
meeting status in a room, can be considered as events. Mostly,
event detection needs to be done in real-time. However, events
such as trends may be detected using historic data. The symbol
() is used to denote that event detection functionality is
facilitated and employed by the project in some perspective.
12) Context Discovery and Annotation: We use the following abbreviations to denote context discovery and annotation facilitated and employed by the project: context
discovery (D) and context annotation (A). Context annotation
allows context related information and raw sensors data to be
attached, modelled, and stored. Some of the most common and
basic information that needs to be captured in relation to context are context type, context value, time stamp, source, and
confidence. Context-aware geographical information retrieval
approach [162] has proposed a mechanism to map raw sensor
data to semantic ontologies using SWRL. This is critical
in all types of systems. Even though, statistical reasoning
systems can use raw sensor data directly, semantic mapping
before the reasoning allows more information to be extracted.
Context information only becomes meaningful when it is
interpreted with respect to the user. This can be achieved by
knowledge base integration and reasoning using ontologies.
Another application is discussed in [161]. Ontologies and other
context modelling techniques allow structure data to be more
meaningful which express relationships among data.
End-users in the IoT paradigm are more interested in highlevel information compared to low-level raw sensor data [50].
The following examples explain the difference between highlevel information and low-level raw sensor data. It is raining
(high-level information) can be derived from humidity is 80%
(low-level sensor data). Further, high-level sensor data can be
explained as semantic information as it provides more meaning
to the end users. Challenges of semantic sensor webs are
identified and discussed in [187]. This is the most common
form of discovery.
13) Level of Context Awareness: Context-awareness can
be employed at two levels: low (hardware) level and high
(software) level. At the hardware level, context-awareness is
used to facilitate tasks such as efficient routing, modelling,
reasoning, storage and event detection (considering energy
consumption and availability) [188]. At the hardware level,
data and knowledge available for decision making is less.
Further, sensors are resource constraint devices, so complex
processing cannot be performed at the hardware level. However, applying context-aware technologies in the hardware
level allows resources to be saved, such as network communication costs by preliminary filtering. The software level has
access to a broader range of data and knowledge as well as
more resources, which enables more complex reasoning to be
performed. We use the following abbreviations to denote the
level of context awareness facilitated and employed by the
project: high level (H) and low level (L).
14) Security and Privacy: This is a major concern in
context-aware computing in all paradigms. However, the IoT
paradigm will intensify the challenges in security and privacy.
In the IoT, sensors are expected to collect more information
about users (i.e. people) in all aspects. This includes both
437
physical and conceptual data, such as location, preferences,

calendar data, and medical information to name a few. As
a result, utmost care needs to be taken when collecting,
modelling, reasoning, and with persistent storage. Security and
privacy need to be handled at different levels in the IoT. At
the lowest level, the hardware layer should ensure security
and privacy during collecting and temporary storage within
the device. Secure protocols need to ensure communication
is well protected. Once the data is received, application level
protection needs to be in placed to monitor and control who
can see or use context and so on. Different projects use
different techniques such as policies, rules, and profiles to
provide security and privacy. The symbol () denoted the
presence of security and privacy related functionality in the
project, in some form.
15) Data Source Support: There are different sources
that are capable of providing context. Broadly we call them
sensors. We discussed different types of sensors in Section III.
Based on the popularity of the data sources supported by each
solution, we selected the following classification. (P) denotes
that the solution supports only physical sensors. Software
sensors (S) denotes that the solution supports either virtual
sensors, logical sensors or both. (A) denotes that the solution
supports all kinds of data sources (i.e. physical, virtual, and
logical). (M) denotes that the solution supports mobile devices.
16) Quality of Context: We denote the presence of conflict
resolution functionality using (C) and context validation functionality using (V). Conflict resolution is critical in the context
management domain [189]. There has to be a consistency in
collecting , aggregating, modelling, and reasoning. In the IoT
paradigm, context may not be accurate. There are two reasons
for context information not to be certain. First is that the sensor
technology is not capable of producing 100% accurate sensor
data due to various technical and environmental challenges.
Secondly, even with sensors that produce 100% accurate
sensor data, reasoning models are not 100% accurate. In
summary, problems in sensor technology and problems in
reasoning techniques contribute to context conflicts. There are
two types of context conflicts that can occurred and they are
defined in [189]:
Internal context conflict: Fusing two or more context elements that characterises the situation from different dimensions of the same observed entity in a given moment may
lead to internal context conflict. (e.g. motion sensor detects
that a user is in the kitchen and calendar shows that the
user is supposed to be in a meeting. Therefore, it is unable
to correctly deduce the current location by fusing two data
sources: calendar and motion sensor.)
External
context
conflicts:
The
context
conflict/inconsistency that may occur between two or more bits
of context that describe the situation of an observed entity
from the same point of view. (e.g. two motion sensors
located in the same area provide two completely different
readings, where one sensor detects a person and other
sensor detects three people.)
Context validation ensures that collected data is correct and
meaningful. Possible validations are checks for range, limit,
logic, data type, cross-system consistency, uniqueness, cardinality, consistency, data source quality, security, and privacy.
438
TABLE XII
S UMMARIZED TAXONOMY USED IN TABLE XIII
Taxonomy
5
Modelling
Reasoning
Distribution
Architecture
9
10
11
12
13
14
15
16
17
18
19
20
History and Storage

Knowledge Management
Event Detection
Context Discovery and Annotation
Level of Context Awareness
Security and Privacy
Data Source Support
Quality of Context
Data Processing
Dynamic Composition
Real Time Processing
Registry Maintenance
Description
Key-value modelling (K), Markup schemes (M), Graphical modelling (G), Object oriented
modelling (Ob), Logic-based modelling (L), and Ontology-based modelling (On)
Supervised learning (S), Un-supervised learning (U), rules (R), Fuzzy logic (F), Ontology-based
(O), and Probabilistic reasoning (P)
Publish/subscribe (P) and Query (Q)
Component based architecture (1) , Distributed architecture (2), Service based architecture (3),
Node based architecture (4) , Centralised architecture (5), Client-server architecture (6)
Available ()
Available ()
Available ()
context Discovery (D) and context Annotation (A)
High level (H) and Low level (L).
Available ()
Physical sensors (P), Software sensors (S), Mobile devices (M), Any type of sensor (A)
Conflict resolution (C), context Validation (V)
Aggregate (A), Filter (F)
Available ()
Available ()
Available ()
17) Data Processing: We denote the presence of context

aggregation functionality using (A) and context filter functionality using (F). Aggregation can be explained in different
ways; for example, Context Toolkit [72] has a dedicated
component called context aggregator to collect data related to a
specific entity (e.g. person) from different context sources and
act as a proxy to context applications. They do not perform any
complex operations; just collect similar information together.
This is one of the simplest forms of aggregation of context.
Context filter functionality makes sure the reasoning engine
processes only important data. Specially in IoT, processing all
the data collected by all the sensors is not possible due to
scale. Therefore, IoT solutions should process only selected
amounts of data that allows it to understand context accurately.
Filtering functionality can be presented in different solutions
in different forms: filter data, filter context sources, or filter
events. Filtering helps both at the low (hardware) level and
software level. At the hardware level, it helps to reduce the
network communication cost by transmitting only important
data. At the high-level, filtering can save process energy by
only processing important data.
Context processing can be classified into three categories
(also called layers) [11]. Typical methods and techniques used
in each layer are also presented as follows:
Activity and context recognition layer: Feature extraction,
classification, clustering, fuzzy rules
Context and representation layer: Conceptual models, logic
programming, ontology based representation and reasoning,
databases and query languages, rule based representation
and reasoning, cased based representation and reasoning,
representing uncertainty, procedural programming
Application and adaptation layer: Rules, query languages,
procedural programming
Data fusion, which is also considered a data processing
technique, is critical in understanding sensor data. In order
to lay a solid foundation to our discussion, we adopt the
definition provided by Hall and Llinas [149] on sensor data
fusion. Sensor data fusion is a method of combining sensor data from multiple sensors to produce more accurate,
more complete, and more dependable information that could

not be possible to achieve through a single sensor [149].
For example, in positioning, GPS does not work indoors.
In contrast, there are a variety of other indoor positioning
schemes that can be used. Therefore, in order to continuously
track the positioning regardless of indoor or outdoor, sensor
data fusion is essential [78]. Data fusion methods, models,
and classification techniques in the wireless sensor networks
domain are comprehensively surveyed in [190].
In order to identify context, it is possible to combine data
from different data sources. For example, consider a situation
where we want to identify the location of a user. The possible
sources that can be used to collect evidence regarding the
location are GPS sensors, motion sensor, calendar, email,
social networking services, chat clients, ambient sound (sound
level, pattern), users nearby, camera sensors, etc. This long
list shows the possible alternatives. It is always a tradeoff
between required resource (e.g. processing power, response
time) and accuracy. Processing and combining all the above
sensor readings would produce a more accurate result; however, it would require more resources and time. There is a
significant gap between low-level sensor readings and highlevel situation-awareness [123]. Collecting low-level sensor
data is becoming significantly easier and cheaper than ever
due to advances in sensing technology. As a result, enormous
amounts of sensor data (e.g. big data [5]) is available. In
order to understand big data, a variety of different reasoning
techniques need to employed as we discussed in Section IV-C.
18) Dynamic Composition: As explained in Solar [184],
IoT solutions must have a programming model that allows
dynamic composition without requiring the developer or user
to identify specific sensors and devices. Dynamic organising is
critical in environments like the IoT, because it is impossible
to identify or plan possible interaction at the development
stage. Software solutions should be able to understand the
requirements and demands on each situation, then organise
and structure its internal components according to them.
Components such as reasoning models, data fusion operators,
knowledge bases, and context discovery components can be
dynamically composed according to the needs. The symbol

() denoted the presence of dynamic composition functionality in the project in some form.
19) Real Time Processing: Most of the interactions are
expected to be processed in real time in the IoT. This functionality has been rarely addressed by the research community
in the context-aware computing domain. The most important
real time processing task is event detection as we explained
in Section V-A11. However, context reasoning, and query
processing can also be considered as essential real time
processing tasks. Real time processing solutions are focused
on processing faster than traditional methods, which allows
sensor stream data processing [211]. The symbol () denoted
the presence of real time processing functionality in some
form.
20) Registry Maintenance and Lookup Services: We use
the () symbol to denote the presence of registry maintenance
and lookup services functionality in the project. This functionality allows different components such as context sources, data
fusion operators, knowledge bases, and context consumers
to be registered. This functionality is also closely related to
dynamic composition where it needs to select relevant and
matching components to be composed together. Registries
need to be updated to reflect (dis)appearing components.
B. Evaluation of Research Efforts
Context Toolkit [72] aims to facilitating development and
deployment of context-aware applications. This is one of the
earliest efforts of providing framework support for contextaware application development. Context Toolkit contains a
combination of features and abstractions to support contextaware application developers. It introduces three main abstractions: context widget (to retrieve data from sensors), context
interpreter (to reason sensor data using different reasoning
techniques), and context aggregator. The research around
Context Toolkit is still active and a number of extensions
have been developed to enhance its context-aware capabilities.
Enactor [217] provides a context decision modelling facility
to the Context Toolkit. Further, the Intelligibility Toolkit [108]
extends the Enactor framework by supporting more decision
models for context reasoning. Context Toolkit identifies the
common features required by context-aware applications as
capture and access of context, storage, distribution, and independent execution from applications.
Aura [191] is a task oriented system based on distributed
architecture which focuses on different computational devices
used by human users every day. The objective is to run a set
of applications called personal aura in all devices in order
to manage user tasks in a context-aware fashion across all
the devices smoothly. Aura addresses two major challenges.
First, aura allows a user to preserve continuity in his/her
work when moving between different environments. Second,
it is capable of adapting to the on-going computation of a
particular environment in the presence of dynamic resource
variability. Aura consists of four major components: context
observer (collects context and send it to task and environment
managers), task manager (also called prism, four different
kinds of changes: user moves to another environment, environment, task, and context), environment manager (handles
439
context suppliers and related service), and context suppliers

(provides context information). XML based markup schemes
are used to describe services. .
CARISMA [193] (Context-Aware Reflective middleware
System for Mobile Applications) is focused on mobile systems
where they are extremely dynamic. Adaptation (also called
reflection) is the main focus of CARISMA. context is stored
as application profiles (XML based), which allows each application to maintain meta-data under two categories: passive and
active. The passive category defines actions that middleware
would take when specific events occur using rules, such as
shutting down if battery is low. However, conflicts could
arise when two profiles defines rules that conflict each other.
The active category allows relationships to be maintained
between services used by the application, the policies, and
context configurations. This information tells how to behave
under different environmental and user conditions. A conflict
resolution mechanism is also introduced in CARISMA based
on macroeconomic techniques. An auction protocol is used
to handle the resolution as they support greater degrees of
heterogeneity over other alternatives. In simple terms, rules
are used in auctions with different constraints imposed on the
bidding by different agents (also called applications). Final
decisions are made in order to maximise the social welfare
among the agents.
CoBrA [119] (Context Broker Architecture) is a brokercentric agent architecture that provides knowledge sharing and
context reasoning for smart spaces. It is specially focused on
smart meeting places. CoBrA addresses two major issues:
supporting resource-limited mobile computing devices and
addressing concerns over user privacy. Context information
is modelled using OWL ontologies. Context brokers are the
main elements of CoBrA. A context broker comprises the
following four functional components: context knowledge base
(provides persistent storage for context information), context
reasoning engine (performs reasoning over context information stored in storage), context acquisition module (retrieve
context from context sources), and policy management module
(manages policies, such as who has access to what data). Even
though the architecture is centralised, several brokers can work
together through a broker federation. Context knowledge is
represented in Resource Description Framework (RDF) triples
using Jena.
Gaia [168] is a distributed context infrastructure uncertainty
based reasoning. Ontologies are used to represented context
information. Gaia has employed a Prolog based probabilistic
reasoning framework. The architecture of Gaia consists of
six key components: context provider (data acquisition from
sensors or other data sources), context consumer (different parties who are interest in context), context synthesiser (generate
high-level context information using raw low-level context),
context provider lookup service (maintains a detailed registry
of context providers so the appropriate context providers can
be found based on their capabilities when required), context
history service (stores history of context), and ontology server
(maintains different ontologies).
SOCAM [194] (Service Oriented Context-Aware Middleware) is an ontology based context-aware middleware. It
separates the ontologies into two levels: upper level ontology
440
Project Name
Citations
Year
Project Focus
Modelling
Reasoning
Distribution
Architecture
History and Storage
Knowledge
Management
Event Detection
Context Discovery
and Annotation
Level of Context
Awareness
Security and Privacy
Data Source Support
Quality of Context
Data Processing
Dynamic Composition
Real Time Processing
Registry Maintenance
TABLE XIII
E VALUATION OF S URVEYED R ESEARCH P ROTOTYPES , S YSTEMS , AND A PPROACHES
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
(9)
(10)
(11)
(12)
(13)
(14)
(15)
(16)
(17)
(18)
(19)
(20)
Context Toolkit
[72]
2001
1,5
Solar
Aura
[184]
[191]
2002 M
2002 M
K,M,Ob
M
R
R
P
P
2
2

D
D
H
H
P
A
CoOL
CARISMA
[192]
[193]
2003 T
2003 M
On
M
R,O
R
Q
Q
1
2
H
H
S
M
CoBrA
Gaia
SOCAM
[119]
[168]
[194]
2004 M
2004 M
2004 M
On
F,On
On
R,O
S,P, F
R,O
Q
Q
Q,P
1
2,3
3

D
D
H
H
H

A
A
A

CARS
CASN
SCK
TRAILBLAZER
[195]
[188]
[142]
[196]
2005 S
2005 M
2005 M
2005 S
K
F,On
M,On
K
U
F,O
R,O
R
P
Q
Q
2
1
2

A
D
A,D
D
H
L
H
L
P
P
A
P
BIONETS
PROCON
CMF (MAGNET)
e-SENSE
[197]
[86]
[85]
[44]
2006 M
2006 S
2006 M
2006 M
On
K
M
R,O
R
R
R
Q
Q
P,Q
Q
1
2
2,4
2,4
A
D
D
D
H
L
H
H
A
P
A
P
A,F
HCoM
CMS
MoCA
CaSP
SIM
[198]
[106]
[199]
[185]
[200]
[124]
2007
2007
2007
2007
2007
2007
M
M
M
M
M
M
G,On
On
M,Ob
M,On
K,G
On
R,O
O
O
O
R
O
Q
P,Q
P,Q
P,Q
5
1,2
4,5
6
2

D
S
D
D
H
H
H
H
H
H
S
A
A
A
P
P
C
V
F
A
A
A

COSMOS
DMS-CA
CDMS
AcoMS
CROCO
EmoCASN
[201]
[202]
[203]
[141]
[204]
[88]
[118]
[205]
2008
2008
2008
2008
2008
2008
2008
2008
M
S
M
M
M
M
M
S
Ob
M
K,M
On
On
M,G,On
On
K
R
R
R
O,P
R,O
R,O
R,O
R
Q
Q
Q
Q
P,Q
P
Q
Q
2,4
5
2
5
5
5
2,4

D
D
D
A
D
D
H
H
H
H
H
H
H
L
P
A
A
P
P
A
P
C,V
A,F

Hydra
UPnP
COSAR
SPBCA
C-CAST
CDA
SALES
MidSen
[61]
[206]
[151]
[161]
[207]
[208]
[209]
[210]
[52]
2009
2009
2009
2009
2009
2009
2009
2009
2009
M
M
M
M
M
M
M
M
M
K,On,Ob
K,M
On
On
M
On
Ob
M
K
R,O
R
S,O
R,O
R
O
R
R
Q
Q
Q
Q
P,Q
P
Q
Q
P,Q
3
4
5
2
5
5
4,6
2,4
5

D
A
A
D
D
D
D
H
H
H
H
H
H
H
L
H

P
A
P
A
A
A
V
P
P

SCONSTREAM
Feel@Home
CoMiHoC
Intelligibility
ezContext
UbiQuSE
COPAL
[211]
[101]
[212]
[213]
[108]
[105]
[214]
[215]
2010
2010
2010
2010
2010
2010
2010
2010
S
M
M
M
T
M
M
M
G
M
G,On
Ob
K,Ob
M
M
R
P
O
R,P
R,S,P
R
R
R
Q
Q
P,Q
Q
Q
Q
Q
P,Q
5
2,4
2,4
5
1,5
5
5
1,5

D
D
D,A
D
H
H
H
H
H
H
H
H
P
A
A
A
A
A
A
V
V
A,F

[50]

P
2,4
Octopus
2011 S
[216]

P
2
2011 M
[153]
K,Ob
S,P
2,4

2011 S
Notes: Refer Section V-A for the meanings of the abbreviations and symbols used in the table
D
D
D,A
H
H
H
A
P
M
A
A
A,F

for general concepts and lower level ontologies domain specific descriptions. SOCAM architecture comprises several key
components: context provider (acquires data from sensors and
other internal and external data sources and converts the context in to OWL representation), context interpreter (performs
reasoning using reasoning engine and stores the processed
context information in the knowledge base), context-aware
services (context consumers), and services locating service
(context providers and interpreter are allowed to register so
other components can search for appropriates providers and
interpreters based on their capabilities).
e-SENSE [44] enables ambient intelligence using wireless
multi-sensor networks for making context-rich information
available to applications and services. e-SENSE combines
body sensor networks (BSN), object sensor networks (OSN),
and environment sensor networks (ESN) to capture context in
the IoT paradigm. The features required by context-aware IoT
middleware solutions are identified as sensor data capturing,
data pre-filtering, context abstraction data source integration,
context extraction, rule engine, and adaptation.
HCoM [198] (Hybrid Context Management) is a hybrid
approach which combines semantic ontology and relational
schemas. This approach claims that standard database management systems alone cannot be used to manage context. In
contrast, semantic ontologies may not perform well in terms
of efficiency and query processing with large volumes of
data. So the hybrid approach is required. HCoM architecture
consists of five layers: acquisition layer, pre-processing layer,
data modelling and storage layer, management modelling
layer, and utilising layer. HCoM has identified several key
requirements that a context management solution should have
that are encapsulated in several components: context manager
(aggregates the results and sends the data to reasoning engine),
collaboration manager (if context selector decides the existing
context information is not sufficient to perform reasoning, the
collaboration manager attempts to gather more data from other
possible context sources), context filter (once the context is
received, it validates and decide whether it needs to be stored
in RCDB), context selector (based on the user request, it
decides what context should be used in reasoning processing
based on the accuracy, time, and required computational
resources), context-onto (manages the ontologies and acts as a
repository), rules and policy (users are allowed to add rules to
the system), RCDB (stores the captured context in a standard
database management system), rule-mining (a data base that
consists of rules that tell what actions to perform when), and
interfaces (provides interface to the context consumers).
MoCA [199] is a service based distributed middleware
that employs ontologies to model and manage context. The
primary conceptual component is context domain. The context
management node (CMN) is infrastructure that is responsible
for managing the context domain. Similar to most of the other
context management solutions, the three key components in
MoCA are: context providers (responsible for generating or
retrieving context from other sources available to be used by
the context management system), context consumers (consume
the context gathered and processed by the system), and context
service (responsible for receiving, storing, and disseminating
context information). MoCA uses an object oriented model for
441
context handling, instead of an ontology-based model due to

the weaknesses posed by ontologies in terms of scalability and
performance. XML is used to model context. The XML files
are fed into the context tool in order to check validation. Then
the program codes are generated automatically to acquire data.
These program codes will acquire context and insert the data
into context repositories.
CaSP [185] (Context-aware Service Platform) is a context gathering framework for mobile solutions based on
middleware architecture. The platform provides six different
functionalities: context sensing, context modelling, context
association, context storage, and retrieval. The paper also
provides a comprehensive evaluation of existing context sensing solutions. CaSP consists of typical context management
components which handle the mentioned functionalities.
SIM [200] (Sensor Information Management) is focused
on the smart home domain which addresses location tracking. SIM uses an agent based architecture according to the
standard specifications provided in Foundation for Intelligent
Physical Agents. Its emphasis is on collecting sensor data from
multiple sources and aggregating them together to analyse
and derive more accurate information. SIM collects two types
of information: node level and attribute level. In node level,
node ID, location, and priority are collected. Attributes are
stored in attribute information base comprising attribute and
the corresponding measurement. A location tracking algorithm
has been introduced using a mobile positioning device. A
position manager handles tracking. SIM has the capability
to resolve conflicts in sensor information based on sensor
priority. Conflict resolution is handled by a context manager
with the help of aggregation, classification, and decision
components. Even though SIM is not focused on hardware
level context management, the approach is closer to low-level
instead of high-level compared to other projects.
COSMOS [201] is middleware that enables the processing
of context information in ubiquitous environments. COSMOS
consists of three layers: context collector (collects information
from the sensors), context processing (derives high level
information from raw sensor data), and context adaptation
(provides access to the processed context for the applications).
In contrast to the other context solutions, the components of
COSMOS are context nodes. In COSMOS, each piece of context information is defined as a context node. COSMOS can
support any number of context nodes which are organised into
hierarchies. Context node is an independently operated module
that consists of its own activity manager, context processor,
context reasoner, context configurator, and message managers.
Therefore, COSMOS follows distributed architecture which
increases the scalability of the middleware.
DMS-CA [202] (Data Management System-Context Architecture) is based on smart building domain. XML is used to
define rules, contexts, and services. Further, an event driven
rule checking technique is used to reason context. Rules can
be configured by mobile devices and push them to the server
to be used by the rule checking engine. Providing a mobile
interface to build rules and queries is important in a dynamic
and mobile environment such as the IoT.
ACoMS [88] (Autonomic Context Management System)
can dynamically configure and reconfigure its context infor-
442
mation acquisition and pre-processing functionality to perform

fault tolerant provisioning of context information. ACoMS
architecture comprises application context subscription manager stores (manages context information requests from the
applications using a subscribe mechanism), context source
manager (performs actions such as low-level communication
with context sources, context source discovery, registration,
and configuration), and reconfiguration manager (performs
monitoring tasks such as mapping context sources to context
information).
CROCO [118] (CROss application COntext management)
is an ontology based context modelling and management
service. CROCO identifies several requirements to be a cross
application, such as application plug-in capability. CROCO
has three responsibilities where they are distributed among
three separate layers: data management (perform operations
such as storing inferred data for historic use, develop and
maintain fact database), consistency checking and reasoning
(consistency manager is responsible for checking the consistency, such as data types, and cardinality when sensor data
arrives before it is feed in to reasoning or storage; reasoning
manager performs reasoning based on the facts stored in the
fact data base), and context update and provision (allows
context consumers to register themselves, retrieve context from
context sources, and provide query interface to the consumers).
EMoCASN [205] (Environment Monitoring Oriented Context Aware Sensor Networks) proposes a context-aware model
for sensor networks (CASN). This modelling approach is
narrowly focused on managing sensor networks using low
level context such as node context, task context, and data
context. For example, CASN uses low level context such
as remaining energy of a node, location of the sensor, and
orientation of the sensor to decide energy efficient routing.
Hydra3 [61] is an IoT middleware that aims to integrate
wireless devices and sensors into ambient intelligence systems.
Hydra comprises a Context Aware Framework (CAF). CAF
provides the capabilities of both high-level, powerful reasoning, based on the use of ontologies and lower-level semantic
processing based on object-oriented/key-value approach. CAF
consists of two main components: Data Acquisition Component (DAqC) and the Context Manager (CM). DAqC is
responsible for connecting and retrieving data from sensors.
CM is responsible for context management, context awareness,
and context interpretation. A rule engine called Drools platform [218] has been employed as the core context reasoning
mechanism. CAF models three distinct types of context:
device contexts (e.g. data source), semantic contexts (e.g.
location, environment, and entity), and application contexts
(e.g. domain specific). Hydra identifies context reasoning rule
engine, context storage, context querying, and event/action
management as the key components of a context-aware framework.
C-Cast [207] is middleware that integrates WSN into
context-aware systems by addressing context acquisition, dissemination, representation, recognising, and reasoning about
context and situations. C-Cast lays its architecture on four
3 The name Hydra has changes its name due to name conflict between
another project registered under same name in Germany. The new name of
the middleware is the LinkSmart middleware.
layers: sensor, context detection, context acquisition, and

application. In C-Cast, context providers (CP) are the main
components. Each context provider handles one task. For example, WeatherCP collects weather information and AddressbookCP collects related addresses. Any amount of CPs can be
added to the system to extend the system wide functionality.
Each context provider independently handles data acquisition,
context processing (e.g. filter and aggregate context), context
provider management (e.g. handles subscriptions), and context
access and dissemination (e.g. handles queries). C-Cast claims
that complex reasoning and intuitive reasoning can only be
achieved by using rich representation models. In contrast,
C-CAST avoids using ontologies to model context claiming
ontologies are too resource intensive.
SALES [210] (Scalable context-Aware middleware for
mobiLe EnviromentS) is a context-aware middleware that
achieves scalability in context dissemination. The main components of this middleware are nodes. These nodes are not
sensor nodes but servers, computers, laptops, PDAs, and
mobile phones. SALES consists of four types of nodes. XML
schemes are used to store and transfer context.
MidSen [52] is context-aware middleware for WSN. The
system is based on Event-Condition-Action (ECA) rules. It
highlights the importance of efficient event detection by processing two algorithms: event detection algorithm (EDA) and
context-aware service discovery algorithm (CASDA). MidSen
has proposed a complete architecture to enable context awareness in WSN. It consists of the following key components:
knowledge manager, application notifiers, knowledge base,
inference engine, working memory, application interface, and
network interface.
Feel@Home [212] is a context management framework that
supports interaction between different domains. The proposed
approach is demonstrated using three domains: smart home,
smart office, and mobile. The context information is stored
using OWL [146]. Feel@Home supports two different interactions: intra-domain and cross domain. The cross domain
interaction is essential in the IoT paradigm. Further, this is
one of the major differences between sensor networks and
the IoT. Sensor networks usually only deal with one domain.
However, IoT demands the capability of dealing with multiple
domains. In addition, context management frameworks should
not be limited to a specific number of domains. Feel@Home
consists of three parts: user queries, global administration
server (GAS), and domain context manager (DCM). User
queries are first received by GAS. It decides what the relevant
domain needs to be contacted to answer the user query.
Then, GAS redirects the user query to the relevant domain
context managers. Two components reside in GAS, context
entry manager (CEM) and context entry engine (CEE), which
performs the above task. DCM consists of typical context
management components such as context wrapper (gathers
context from sensors and other sources), context aggregator
(triggers context reasoning), context reasoning, knowledge
base (stores context), and several other components to manage
user queries, publish/subscribe mechanism. The answers to
the user query will return by using the same path as when
received.
CoMiHoc [213] (Context Middleware for ad-HoC network)
is a middleware framework that supports context management and situation reasoning. CoMiHoc proposes a CoMoS
(Context Mobile Spaces), a context modeling, and situation
reasoning mechanism that extends the context spaces [219].
CoMiHoc uses Java Dempster-Shafer library [220]. CoMiHoc
architecture comprises six components: context provisioner,
request manager, situation reasoner, location reasoner, communication manager, and On-Demand Multicast Routing Protocol
(ODMRP).
ezContext [105] is a framework that provides automatic
context life cycle management. ezContext comprises several
components: context source (any source that provides context,
either physical sensors, databases or web service), context
provider (retrieves context from various sources whether in
push (passive) or pull (active) method, context manager
(manages context modelling, storage and producing high-level
context using low-level context), context wrapper (encapsulate
retrieved context into correct format, in this approach, keyvalue pairs), and providers registry (maintains list of context
providers and their capabilities). JavaBeans are used as the
main data format.
Octopus [50] is an open-source, dynamically extensible
system that supports data management and fusion for IoT
applications. Octopus develops middleware abstractions and
programming models for the IoT. It enables non-specialised
developers to deploy sensors and applications without detailed
knowledge of the underlying technologies and network. Octopus is focused on the smart home/office domain and its
main component is solver. Solver is a module that performs
sensor data fusion operations. Solvers can be added and
removed from the system at any time based on requirements.
Further solvers can be combined together dynamically to build
complex operations.
VI. L ESSONS L EARNED
1) Development Aids and Practices: Toolkits in general
are suitable for limited scale application. Managing context
in the IoT paradigm requires middleware solutions that can
provide more functionality towards managing data. Applications should be able to be built on top of the middleware
so they can request context from the middleware. Context
Toolkit [72] has introduced the notion of having common
standard interfaces. For example, context widget component
encapsulate the communication between context sources and
the toolkit. Standardisation makes it easier to learn, use, and
extend the toolkit. Standardisation is important in the IoT
paradigm, because it increases interoperability and extendibility. For example, standardising context modelling components
will help to employ the different techniques we discussed in
Section IV-B despite the differences in inner-workings. It also
enables the addition of different components when necessary.
In such a situation, standard interfaces and structures will guarantee a smooth interaction between new and old components.
Further, Intelligibility Toolkit [108] provides explanations to
the users to improve the trust between users and the contextaware applications which helps in faster adaptation of the users
towards IoT.
Making correct design decisions is a critical task in IoT.
For example, data modelling and communication can be done
443
using different techniques as follows where each method has

its own advantages and disadvantages [184]. 1) Binary is
smaller in size than the other three formats and also portable
due to its small size. In contrast, binary makes it difficult to
extend and modify later. 2) Objects method allows complex
data structures. 3) Attribute-value pairs method provides more
limited complexity than an object representation. In contrast,
simpler representation allows language- and platform- independent applications. 4) XML method provides more opportunities for complex data structures. XML adds a substantial
overhead in term of network communications and processing.
CoOL [192] shows how extensions (e.g. context modelling
and reasoning) can be developed to support general purpose
service models. CoOL allows context management functionality to be added to any model using context management
access point, which is responsible for handle communication
between CoOL and the rest of the general purpose architecture.
Security and privacy issues in context-aware computing are
not researched and seriously considered in many solutions.
CoBrA [119] shows how an ontology based approach can be
used to manage user privacy via policies which allow it to
monitor and access contextual control context. As ontologies
are getting popular and adopted in web related developments,
such practice will makes IoT development much easier.
Octopus [50] highlights the significance of designing programming models that enable non-technical people to deploy
sensors. As we mentioned earlier, the majority of the sensor
deployments are expected to be carried out by non-technical
users. Kim and Choi [204] models context meta-data from an
operational perspective as discussed in Section III-C, which
allows it to understand operational parameters such as complexities, quality, up-to-dateness, and cost of acquisition.
2) Mobility, Validity, and Sharing: Monitoring continuity,
which is also called mobility, is an important task in the IoT.
People move from one situation to another and IoT solutions
need to track user movements and facilitate context-aware
functionalities over different forms of devices. Aura [191]
shows the requirement of having IoT middleware running
over many platforms and devices under different resource
limitations (i.e. from cloud server, computers, tablets, mobile
phones to everyday objects) where different versions (with
different capabilities) would fit on different devices.
CARISMA [193] shows how conflict resolution can be
done using profiles and rules where it stress the importance
of making decisions to optimize the return for every party
involved. In the IoT, there will be many data sources that
will provide similar information that can be used to derive the
same knowledge where conflict resolution will help to make
accurate actions. MoCA [199] also emphasizes validation of
context which has an impact on the accuracy of the reasoning.
Further, it shows how context can be modelled in formats such
as XML and then inserted into any programming language via
binding techniques (e.g. data binding in Java). In CROCO
[118], validation (e.g. consistency), conflict resolution, and
privacy concerns are given attention where they are rarely addressed by many other solutions. Sharing context information
allows mobility and smooth transition from device to device or
situation to situation. Park et al. [153] highlight the importance
of context sharing using mobile devices, which allows more
444
comprehensive and accurate reasoning and high level context

recognition.
3) On Demand Data Modelling: Due to unpredictability
and broadness of IoT, data models need to be extensible on
demand. For example, IoT solutions may need to be expand
its knowledge-base towards different domains. SOCAM [194]
shows how knowledge can be separated among different levels
of ontologies: upper ontology and domain specific ontology.
In SOCAM, upper ontology models general purpose data
while domain specific ontologies model domain specific data,
which is allowed to extend to both levels independently. As
an IoT solution will be used in many different domains, the
ability to add ontologies (i.e. knowledge) when necessary is
critical for wider adaptation. SCK [142], Zhan et al. [208],
and BIONETS [197] use different ontologies for each context
category. As we discussed in Section III-C, there are many
different types of context categories which model context in
different perspectives. Therefore, in the IoT it is important to
store different types of context as they can help in a variety
of situations. They also stresses the requirement of having
domain specific and domain independent ontologies.
4) Hybrid Reasoning: Gaia [168], Ko and Sim [141],
CDMS [203], and HCoM [198] highlight the importance of
employing multiple reasoning techniques such as Bayesian
networks, probabilistic and fuzzy logic, where each technique performs well in different situations. Incorporation of
multiple modelling and reasoning techniques can mitigate
individual weaknesses using each others strengths. COSAR
[151] combines statistical reasoning and ontological reasoning
techniques to achieve more accurate results.
5) Hardware Layer Support: EMoCASN [205], TRAILBLAZER [196] and CASN [188] shows the importance of
embedding context-aware capabilities in low (hardware) layer
communication. Context awareness allows sensors to act more
intelligently and save energy. In the IoT a majority of the
communications are expected to happened between machines.
In such situations, context awareness becomes critical for each
individual object to optimize their actions. Further, in order
to build a fully context-aware solution, we have to embed
context-aware capabilities in both software and hardware
layers. In an environment such as the IoT where billions
of objects communicate with each other, significant amounts
of energy can be saved by following fairly simple optimisation techniques as presented in PROCON [86]. SALES
[210] shows how context can be managed using distributed
architecture with a variety of different devices with different
resource constraints in the hardware level.
6) Dynamic Configuration and Extensions: Hydra [61] is
one of the early efforts at building IoT middleware which
focuses on connecting embedded devices to applications.
It shows how the context modelling needs to be done in
order to model device information. Hydra also highlights
the importance of pluggable rules that allow insertions when
necessary as it is a major requirement in IoT middleware
applications, where domains and required knowledge cannot
be predicted during the development stage. A complementary
technology has proposed by ACoMS [88]. It has proposed a
technique that allows it to automatically connect sensors to an
IoT solution using Transducer Electronic Data Sheet (TEDS)
[221] and Sensor Markup Languages (SensorML) [133]. UPnP

FRAMEWORK [206] is strongly related to a vision of the IoT
where machine-to-machine communication play a significant
role. This approach is applicable to devices such as cameras,
web cams, and microwaves; but, not for low end temperature
or humidity sensors. UPnP approach is a key technology that
enables automated configuration.
Solar [184], CMF (MAGNET) [85], and COSMOS [201]
promote the notion of dynamic composition which is critical in
IoT solutions where possible interactions cannot be identified
at the design and development stage. ezContext [105] shares
a common notion of context providers similar to C-Cast
[207] uses them to decouple context sources from the system.
Different types of context providers, which are dedicated
to communicating and retrieving data related to a specific
domain, can be employed when necessary. In line with above
solutions, COPAL [215] demonstrates the essential features
IoT middleware should have, such as loosely coupled plugin architecture and automated code generation via abstracts
which stimulates extendibility and usability.
7) Distributed Processing: This is a one of the most
commons tasks need to be performed by IoT solutions.
UbiQuSE [214] shows how real time query processing can
be done incorporating live streaming data and historic context
in repositories. Similarly, SCONSTREAM [211] highlights the
challenges in real-time context stream processing where real
time processing is a significant component to be successful
in the IoT. Most event detections need to be performed
in real time. Further, Feel@Home [212] shows how cross
domain context can be queried in order to answer complex
user requirements. As we mentioned earlier, there is not a
central point of management in the IoT paradigm. Therefore
communicating, sharing, and querying context managed in a
distributed fashion by different managers is essential.
8) Other Aspects: CARS [195] introduces a technique that
can be used to evaluate, test, and improve IoT solutions in
social and user point of view. As we mentioned earlier, success
of IoT depends on the user adaptation. CARS evaluates the
process of deriving high level information using low level
sensor data where users will appreciate the work done by the
software systems.
Cloud computing offers significant amounts of processing and storage capabilities. With the three services models, Infrastructure-as-a-Service (IaaS), Platform-as-a-Service
(PaaS), and Software-as-a-Service (SaaS), context management can largely benefit from cloud computing in many ways
in the IoT paradigm. In the IoT, sensors will be attached to
almost every object around us. Further, these sensors will be
deployed by ordinary users, governments, or business organisations. Cloud computing allows all parties to share sensor
data based on a financial model. Sensor owners will advertise
their sensors in the cloud. The consumers who want to access
those sensors will pay the owners and acquire the sensor
readings. Therefore, the cloud model perfectly matches with
the IoT paradigm. In addition, cloud resources can be used to
reason and store large volumes of context where significant
amounts of processing power and storage are required. The
cloud brings added scalability to context management in the
IoT. Further, interoperability among different IoT solution can
be achieved by following approaches such as CDA [209].

Data matching is the process to identify and matching records
in diverse database that refer to the same real-world entities
in situations where no entity identifiers are available, and
therefore the available attributes have to be used to conduct
the matching [222]. Context information plays a critical in
data matching where sensors can be considered as entities in
the sensing as a service model [5].
VII. C HALLENGES AND F UTURE R ESEARCH D IRECTIONS
As we mentioned earlier, one of our goal in this survey is
to understand how context-aware computing can be applied
in the IoT paradigm based on past experience. Specifically,
we evaluated fifty context-aware projects and highlighted the
lessons we can learn from them in the IoT perspective. In this
section our objective is to discuss six unique challenges in
the IoT where novel techniques and solution may need to be
employed.
1) Automated configuration of sensors: In traditional pervasive/ubiquitous computing, we connect only a limited number
of sensors to the applications (e.g. smart farm, smart home). In
contrast, the IoT envisions billions of sensors to be connected
together over the Internet. As a result, a unique challenge
would arise on connection and configuration of sensors to
applications. Due to the scale, it is not feasible to connect
sensors manually to an application or to a middleware [223].
There has to be an automated or at least semi-automated process to connect sensors to applications. In order to accomplish
this task, applications should be able to understand the sensors
(e.g. sensors capabilities, data structures they produce, hardware/driver level configuration details). Recent developments
such as Transducer Electronic Data Sheet (TEDS) [221], Open
Geospatial Consortium (OGC) Sensor Web Enablement related standards such as Sensor Markup Languages (SensorML)
[133], sensor ontologies [143], and immature but promising
efforts such as Sensor Device Definitions [224] show future
directions to carry out the research work further, in order to
tackle this challenge.
2) Context discovery: Once we connect sensors to a software solution, as mentioned above, there has to be a method
to understand the sensor data produced by the sensors and
the related context automatically. We discussed context categorisation techniques comprehensively in Section III-C. There
are many types of context that can be used to enrich sensor
data. However, understanding sensor data and appropriately
annotating it automatically in a paradigm such as the IoT,
where application domains vary widely, is a challenging task.
Recent developments in semantic technologies [135], [143],
[225] and linked data [226], [227] show future directions
to carry out further research work. Semantic technology is
popularly used to encode domain knowledge.
3) Acquisition, modelling, reasoning, and distribution: After analysing acquisition, modelling, and reasoning in different
perspectives, it is evident that no single technique would serve
the requirements of the IoT. Incorporating and integrating
multiple techniques has shown promising success in the field.
Some of the early work such as [12], [183] have discussed
the process in detail. However, due to the immaturity of the
field of IoT, it is difficult to predict when and where to
445
employ each technique. Therefore, it is important to define

and follow a standard specification so different techniques can
be added to the solutions without significant effort. Several
design principles have been proposed by [72], [108] as a step
towards standardisation of components and techniques. The
inner-workings of each technique can be different from one
solution to another. However, common standard interfaces will
insure the interoperability among techniques.
4) Selection of sensors in sensing-as-a-service model: This
is going to be one of the toughest challenges in the IoT.
It is clear that we are going to have access to billions of
sensors. In such an environment, there could be many different
alternative sensors to be used. For example, let us consider a
situation where an environmental scientist wants to measure
environmental pollution in New York city. There are two
main problems: (1) what sensors provide information about
pollution? [228] (2) when there are multiple sensors that can
measure the same parameter (e.g. pH concentration in a lake),
what sensor should be used? [229] In order to answer question (1), domain knowledge needs to be incorporate with the
IoT solution. Manually selecting the sensors that will provide
information about environmental pollution is not feasible in
the IoT due to its scale. In order to answer question (2),
quality frameworks need to be defined and employed. Such a
framework should be able to rank the sensors based on factors
such as accuracy, relevancy, user feedback, reliability, cost, and
completeness. Similar challenges have been addressed in the
web service domain during the last decade [230], [231] where
we can learn from those efforts.
5) Security, privacy, and trust: This has been a challenge
for context-aware computing since the beginning. The advantage of context is that it provides more meaningful information
that will help us understand a situation or data. At the same
time, it increases the security threats due to possible misuse
of the context (e.g. identity, location, activity, and behaviour).
However, the IoT will increase this challenge significantly.
Even though security and privacy issues are addressed at the
context-aware application level, it is largely unattended at
the context-aware middleware level. In the IoT, security and
privacy need to be protected in several layers: sensor hardware
layer, sensor data communication (protocol) layer, context
annotation and context discovery layer, context modelling
layer, and the context distribution layer. IoT is a community
based approach where the acceptance of the users (e.g. general
public) is essential. Therefore, security and privacy protection
requirements need to be carefully addressed in order to win
the trust of the users.
6) Context Sharing: This is largely neglected in the
context-aware middleware domain. Most of the middleware
solutions or architectures are designed to facilitate applications
in isolated factions. Inter-middleware communication is not
considered to be a critical requirement. However, in the
IoT, there would be no central point of control. Different
middleware solutions developed by different parties will be
employed to connect to sensors, collect, model, and reason
context. Therefore, sharing context information between different kinds of middleware solutions or different instances
of the same middleware solution is important. Sensor data
stream processing middleware solutions such as GSN [67]

446

%"-

.
#
$

,
,
+"
+

)!

)#

#

&

%!
!
$
2
+0
#

,
*
#

!

"

#

$

%

+

##

+)#

)#

#

(
+)

(")*
(
'*"
*
'

# !

#
/"

.

%
&
0
#
"
+)#
1

! "#
$
"
$
%&
%
'
'
%
%

#
+

,#

#

#

"

*
"

!
*#
*

)
#
)

#

)

+
"
(

0

)
!
%!'
2#

"#

1

)#
3
3
#
3

"
"
"
"

)

"*"

!
"
(
#)

3)
3
#
(

#

) 3

%

#
"
*! %

"
( %

2
.
!
+

%
#
"

* # *"

'!

"

##

+
## #

"

-

+"

0

0

.""#*

.""

!
''

+"
+
)*

,

#

#
##
1+

!
2 #

#

3
+
4*! ##

!
#

)

Fig. 8. Taxonomy (functionalities commonly supported in existing research prototypes and systems); Conceptual Framework (value added features that need
to be supported by ideal context-aware IoT middleware solution)
have employed this capability to share sensor data among

different instances (e.g. installed and configured in different
computers and locations) where context is not the focus.
However, in contrast to sensor data, context information has
strong relationships between each other (e.g. context modelled
using RDF). Therefore, relationship models also need to
be transferred and shared among different solutions, which
enables the receiver to understand and model the context
accurately at the receivers end.
VIII. C ONCLUSIONS
The IoT has gained significant attention over the last few
years. With the advances in sensor hardware technology and
cheap materials, sensors are expected to be attached to all
the objects around us, so these can communicate with each
other with minimum human intervention. Understanding sensor data is one of the main challenges that the IoT would
face. This vision has been supported and heavily invested
by governments, interest groups, companies, and research
institutes. For example, context awareness has been identified
as an important IoT research need by the Cluster of European
Research Projects on the IoT (CERP-IoT) [21] funded by
the European Union. The EU has allocated a time frame
for research and development into context-aware computing
focused on the IoT to be carried out during 2015-2020.
In this survey paper, we analysed and evaluated contextaware computing research efforts to understand how the
challenges in the field of context-aware computing have been
tackled in desktop, web, mobile, sensor networks, and pervasive computing paradigms. A large number of solutions exist
in terms of systems, middleware, applications, techniques, and
models proposed by researchers to solve different challenges
in context-aware computing. We also discussed some of the
trends in the field that were identified during the survey. The
results clearly show the importance of context awareness in
the IoT paradigm. Our ultimate goal is to build a foundation
that helps us to understand what has happened in the past so
we can plan for the future more efficiently and effectively.
ACKNOWLEDGMENT
Authors acknowledge support from SSN TCP, CSIRO,
Australia and ICT OpenIoT Project, which is co-funded by
the European Commission under seventh framework program,
contract number FP7-ICT-2011-7-287305-OpenIoT. The Author(s) acknowledge help and contributions from The Australian National University.
R EFERENCES
[1] M. Weiser, The computer for the 21st century, Scientific American,
vol. 265, no. 3, pp. 6675, July 1991. [Online]. Available:
http://doi.acm.org/10.1145/329124.329126
[2] B. Schilit and M. Theimer, Disseminating active map information
to mobile hosts, Network, IEEE, vol. 8, no. 5, pp. 22 32, sep/oct
1994. [Online]. Available: http://dx.doi.org/10.1109/65.313011
[3] G. D. Abowd, A. K. Dey, P. J. Brown, N. Davies, M. Smith,
and P. Steggles, Towards a better understanding of context
and context-awareness, in Proc. 1st international symposium on
Handheld and Ubiquitous Computing, ser. HUC 99. London,
UK: Springer-Verlag, 1999, pp. 304307. [Online]. Available:
http://dl.acm.org/citation.cfm?id=647985.743843
447
[4] H. Sundmaeker, P. Guillemin, P. Friess, and S. Woelffle, Vision and challenges for realising the internet of things, European
Commission Information Society and Media, Tech. Rep., March
2010, http://www.internet-of-things-research.eu/pdf/IoT Clusterbook
March 2010.pdf [Accessed on: 2011-10-10].
[5] A. Zaslavsky, C. Perera, and D. Georgakopoulos, Sensing as a service
and big data, in International Conference on Advances in Cloud
Computing (ACC-2012), Bangalore, India, July 2012, pp. 2129.
[6] G. Chen and D. Kotz, A survey of context-aware mobile computing research, Department of Computer Science, Dartmouth College,
Hanover, NH, USA, Tech. Rep., 2000, http://www.cs.dartmouth.edu/
reports/TR2000-381.pdf [Accessed on: 2011-12-05].
[7] T. Strang and C. Linnhoff-Popien, A context modeling
survey, in In: Workshop on Advanced Context Modelling,
Reasoning and Management, UbiComp 2004 - The Sixth
International Conference on Ubiquitous Computing, Nottingham/England, 2004. [Online]. Available: http://elib.dlr.de/7444/1/
Ubicomp2004ContextWSCameraReadyVersion.pdf
[8] M. M. Molla and S. I. Ahamed, A survey of middleware for sensor
network and challenges, in Proc. 2006 International Conference
Workshops on Parallel Processing, ser. ICPPW 06. Washington,
DC, USA: IEEE Computer Society, 2006, pp. 223228. [Online].
Available: http://dx.doi.org/10.1109/ICPPW.2006.18
[9] K. E. Kjaer, A survey of context-aware middleware, in Proc.
25th conference on IASTED International Multi-Conference: Software
Engineering. ACTA Press, 2007, pp. 148155. [Online]. Available:
http://dl.acm.org/citation.cfm?id=1332044.1332069
[10] M. Baldauf, S. Dustdar, and F. Rosenberg, A survey on
context aware systems, Int. J. Ad Hoc Ubiquitous Comput.,
vol. 2, no. 4, pp. 263277, Jun. 2007. [Online]. Available:
http://dx.doi.org/10.1504/IJAHUC.2007.014070
[11] M. Perttunen, J. Riekki, and O. Lassila, Context representation and
reasoning in pervasive computing: a review, International Journal
of Multimedia and Ubiquitous Engineering, vol. 4, no. 4, pp. 128,
2009. [Online]. Available: http://www.sersc.org/journals/IJMUE/vol4
no4 2009/1.pdf
[12] C. Bettini, O. Brdiczka, K. Henricksen, J. Indulska, D. Nicklas,
A. Ranganathan, and D. Riboni, A survey of context modelling and
reasoning techniques, Pervasive Mob. Comput., vol. 6, pp. 161180,
April 2010. [Online]. Available: http://dx.doi.org/10.1016/j.pmcj.2009.
06.002
[13] A. Saeed and T. Waheed, An extensive survey of context-aware
middleware architectures, in Electro/Information Technology (EIT),
2010 IEEE International Conference on, may 2010, pp. 1 6.
[Online]. Available: http://dx.doi.org/10.1109/EIT.2010.5612118
[14] S. Bandyopadhyay, M. Sengupta, S. Maiti, and S. Dutta, Role of
middleware for internet of things, International Journal of Computer
Science and Engineering Survey, vol. 2, pp. 94105, 2011. [Online].
Available: http://airccse.org/journal/ijcses/papers/0811cses07.pdf
[15] P. Makris, D. Skoutas, and C. Skianis, A survey on context-aware
mobile and wireless networking: On networking and computing
environments integration, IEEE Commun. Surveys Tutorials, vol. PP,
no. 99, pp. 1 25, 2012. [Online]. Available: http://dx.doi.org/10.
1109/SURV.2012.040912.00180
[16] P. Bellavista, A. Corradi, M. Fanelli, and L. Foschini, A
survey of context data distribution for mobile ubiquitous systems,
ACM Computing Surveys, vol. xx, no. xx, p. 49, 2013.
[Online]. Available: http://www-lia.deis.unibo.it/Staff/LucaFoschini/
pdfDocs/context survey CSUR.pdf
[17] N. Olifer and V. Olifer, Computer Networks: Principles, Technologies
and Protocols for Network Design.
John Wiley & Sons,
2005. [Online]. Available: http://au.wiley.com/WileyCDA/WileyTitle/
productCd-EHEP000983.html
[18] D. Guinard, Towards the web of things: Web mashups for embedded
devices, in In MEM 2009 in Proc. WWW 2009. ACM, 2009.
[19] Casaleggio Associati, The evolution of internet of things, Casaleggio
Associati, Tech. Rep., February 2011, http://www.casaleggio.
it/pubblicazioni/Focus internet of things v1.81%20-%20eng.pdf
[Accessed on: 2011-06-08].
[20] European Commission, Internet of things in 2020 road map for
the future, Working Group RFID of the ETP EPOSS, Tech.
Rep., May 2008, http://ec.europa.eu/information society/policy/rfid/
documents/iotprague2009.pdf [Accessed on: 2011-06-12].
[21] P. Guillemin and P. Friess, Internet of things strategic research
roadmap, The Cluster of European Research Projects, Tech. Rep.,
September 2009, http://www.internet-of-things-research.eu/pdf/IoT
Cluster Strategic Research Agenda 2009.pdf [Accessed on: 2011-0815].
448
[22] Carnot Institutes, Smart networked objects and internet of

things, Carnot Institutes Information Communication Technologies and Micro Nano Technologies alliance, White Paper, January
2011, http://www.internet-of-things-research.eu/pdf/IoT Clusterbook
March 2010.pdf [Accessed on:2011-11-28].
[23] L. Atzori, A. Iera, and G. Morabito, The internet of things: A
survey, Comput. Netw., vol. 54, no. 15, pp. 27872805, Oct. 2010.
[Online]. Available: http://dx.doi.org/10.1016/j.comnet.2010.05.010
[24] G. Kortuem, F. Kawsar, D. Fitton, and V. Sundramoorthy, Smart
objects as building blocks for the internet of things, IEEE Internet
Computing, vol. 14, no. 1, pp. 44 51, jan.-feb. 2010. [Online].
Available: http://dx.doi.org/10.1109/MIC.2009.143
[25] D. Le-Phuoc, A. Polleres, M. Hauswirth, G. Tummarello, and
C. Morbidoni, Rapid prototyping of semantic mash-ups through
semantic web pipes, in Proc. 18th international conference on World
wide web, ser. WWW 2009. ACM, 2009, pp. 581590. [Online].
Available: http://dx.doi.org/10.1145/1526709.1526788
[26] A. Dohr, R. Modre-Opsrian, M. Drobics, D. Hayn, and G. Schreier,
The internet of things for ambient assisted living, in Information
Technology: New Generations (ITNG), 2010 Seventh International
Conference on, 2010, pp. 804809. [Online]. Available: http:
//dx.doi.org/10.1109/ITNG.2010.104
[27] K. Ashton, That internet of things thing in the real world, things matter more than ideas, RFID Journal, June 2009, http://www.rfidjournal.
com/article/print/4986 [Accessed on: 2012-07-30].
[28] D. L. Brock, The electronic product code (epc) a naming scheme for
physical objects, Auto-ID Center, White Paper, January 2001, http:
//www.autoidlabs.org/uploads/media/MIT-AUTOID-WH-002.pdf [Accessed on: 2011-08-25].
[29] International Telecommunication Union, Itu internet reports 2005: The
internet of things, International Telecommunication Union, Workshop
Report, November 2005, http://www.itu.int/dms pub/itu-s/opb/pol/SPOL-IR.IT-2005-SUM-PDF-E.pdf [Accessed on: 2011-12-12].
[30] T. Lu and W. Neng, Future internet: The internet of things, in
3rd International Conference on Advanced Computer Theory and
Engineering(ICACTE), vol. 5, August 2010, pp. V5376V5380.
[Online]. Available: http://dx.doi.org/10.1109/ICACTE.2010.5579543
[31] L. W. F. Chaves and C. Decker, A survey on organic smart labels for
the internet-of-things, in Networked Sensing Systems (INSS), 2010
Seventh International Conference on, 2010, pp. 161164. [Online].
Available: http://dx.doi.org/10.1109/INSS.2010.5573467
[32] Y. Chen, J. Guo, and X. Hu, The research of internet of
things supporting technologies which face the logistics industry, in
Computational Intelligence and Security (CIS), 2010 International
Conference on, 2010, pp. 659663. [Online]. Available: http:
//dx.doi.org/10.1109/CIS.2010.148
[33] Y.-W. Wang, H.-L. Yu, and Y. Li, Internet of things technology applied
in medical information, in Consumer Electronics, Communications
and Networks (CECNet), 2011 International Conference on, april
2011, pp. 430 433. [Online]. Available: http://dx.doi.org/10.1109/
CECNET.2011.5768647
[34] G. Chong, L. Zhihao, and Y. Yifeng, The research and implement
of smart home system based on internet of things, in Electronics,
Communications and Control (ICECC), 2011 International Conference
on, sept. 2011, pp. 2944 2947. [Online]. Available: http://dx.doi.org/
10.1109/ICECC.2011.6066672
[35] J. Burrell, T. Brooke, and R. Beckwith, Vineyard computing: sensor
networks in agricultural production, IEEE Pervasive Computing,
vol. 3, no. 1, pp. 38 45, jan.-march 2004. [Online]. Available:
http://dx.doi.org/10.1109/MPRV.2004.1269130
[36] L. Lin, Application of the internet of thing in green agricultural
products supply chain management, in Intelligent Computation
Technology and Automation (ICICTA), 2011 International Conference
on, vol. 1, 2011, pp. 10221025. [Online]. Available: http:
//dx.doi.org/10.1109/ICICTA.2011.256
[37] A. Asin and D. Gascon, 50 sensor applications for a smarter
world, Libelium Comunicaciones Distribuidas, Tech. Rep., 2012, http:
//www.libelium.com/top 50 iot sensor applications ranking/pdf [Accessed on: 2012-05-02].
[38] BCC Research, Sensors: Technologies and global markets, BCC
Research, Market Forecasting, March 2011, http://www.bccresearch.
com/report/sensors-technologies-markets-ias006d.html [Accessed on:
2012-01-05].
[39] I. Akyildiz, W. Su, Y. Sankarasubramaniam, and E. Cayirci,
A survey on sensor networks, IEEE Commun. Mag., vol. 40,
no. 8, pp. 102 114, aug 2002. [Online]. Available: http:
//dx.doi.org/10.1109/MCOM.2002.1024422
[40] A. Mainwaring, D. Culler, J. Polastre, R. Szewczyk, and J. Anderson,

Wireless sensor networks for habitat monitoring, in Proc. 1st ACM
international workshop on Wireless sensor networks and applications,
ser. WSNA 02. New York, NY, USA: ACM, 2002, pp. 8897.
[Online]. Available: http://doi.acm.org/10.1145/570738.570751
[41] D. Malan, T. Fulford-jones, M. Welsh, and S. Moulton, Codeblue:
An ad hoc sensor network infrastructure for emergency medical care,
in In International Workshop on Wearable and Implantable Body
Sensor Networks, 2004. [Online]. Available: http://www.eecs.harvard.
edu/mdw/papers/codeblue-bsn04.pdf
[42] S. Rooney, D. Bauer, and P. Scotton, Techniques for integrating
sensors into the enterprise network, IEEE Trans. Network and Service
Management, vol. 3, no. 1, pp. 43 52, jan. 2006. [Online]. Available:
http://dx.doi.org/10.1109/TNSM.2006.4798306
[43] A. R. Da Rocha, F. C. Delicato, J. N. de Souza, D. G.
Gomes, and L. Pirmez, A semantic middleware for autonomic
wireless sensor networks, in Proc. 2009 Workshop on Middleware
for Ubiquitous and Pervasive Systems, ser. WMUPS 09. New
York, NY, USA: ACM, 2009, pp. 1925. [Online]. Available:
http://doi.acm.org/10.1145/1551693.1551697
[44] A. Gluhak and W. Schott, A wsn system architecture to capture
context information for beyond 3g communication systems, in
Intelligent Sensors, Sensor Networks and Information, 2007. ISSNIP
2007. 3rd International Conference on, dec. 2007, pp. 49 54.
[Online]. Available: http://dx.doi.org/10.1109/ISSNIP.2007.4496818
[45] F. Neto and C. Ribeiro, Dynamic change of services in wireless
sensor network middleware based on semantic technologies, in
Autonomic and Autonomous Systems (ICAS), 2010 Sixth International
Conference on, march 2010, pp. 58 63. [Online]. Available:
http://dx.doi.org/10.1109/ICAS.2010.17
[46] K. Al Nuaimi, M. Al Nuaimi, N. Mohamed, I. Jawhar, and K. Shuaib,
Web-based wireless sensor networks: a survey of architectures and
applications, in Proc. 6th International Conference on Ubiquitous
Information Management and Communication, ser. ICUIMC 12.
New York, NY, USA: ACM, 2012, pp. 113:1113:9. [Online].
Available: http://doi.acm.org/10.1145/2184751.2184881
[47] S.-H. Sho, K.-S. Kim, W.-r. Jun, J.-S. Kim, S.-H. Kim, and J.-D. Lee,
Ttcg: three-tier context gathering technique for mobile devices, in
Proc. 5th international conference on Pervasive services, ser. ICPS
08. New York, NY, USA: ACM, 2008, pp. 157162. [Online].
[48] C. Alcaraz, P. Najera, J. Lopez, and R. Roman, Wireless sensor networks and the internet of things: Do we need a complete integration?
in 1st International Workshop on the Security of the Internet of Things
(SecIoT10), 2010.
[49] E. Elnahrawy and B. Nath, Context-aware sensors, in Wireless
Sensor Networks, ser. Lecture Notes in Computer Science, H. Karl,
A. Wolisz, and A. Willig, Eds. Springer Berlin / Heidelberg, 2004, vol.
2920, pp. 7793. [Online]. Available: http://dx.doi.org/10.1007/978-3540-24606-0 6
[50] B. Firner, R. S. Moore, R. Howard, R. P. Martin, and Y. Zhang, Poster:
Smart buildings, sensor networks, and the internet of things, in
Proc. 9th ACM Conference on Embedded Networked Sensor Systems,
ser. SenSys 11. New York, NY, USA: ACM, 2011, pp. 337338.
[Online]. Available: http://doi.acm.org/10.1145/2070942.2070978
[51] S. Alam, M. M. R. Chowdhury, and J. Noll, Senaas: An event-driven
sensor virtualization approach for internet of things cloud, in
Networked Embedded Systems for Enterprise Applications (NESEA),
2010 IEEE International Conference on, November 2010, pp. 16.
[Online]. Available: http://dx.doi.org/10.1109/NESEA.2010.5678060
[52] P. Patel, S. Jardosh, S. Chaudhary, and P. Ranjan, Context aware
middleware architecture for wireless sensor network, in Services
Computing, 2009. SCC 09. IEEE International Conference on, sept.
SCC.2009.49
[53] D. Le-Phuoc, J. X. Parreira, M. Hausenblas, Y. Han, and
M. Hauswirth, Live linked open sensor database, in Proc. 6th
International Conference on Semantic Systems, ser. I-SEMANTICS
10. New York, NY, USA: ACM, 2010, pp. 46:146:4. [Online].
[54] M. Botts, G. Percivall, C. Reed, and J. Davidson, Ogc sensor web
enablement: Overview and high level architecture, in Geosensor
Networks Lecture Notes In Computer Science, S. Nittel, A. Labrinidis,
and A. Stefanidis, Eds. Berlin, Heidelberg: Springer-Verlag, 2008, ch.
OGC Sensor Web Enablement: Overview and High Level Architecture,
pp. 175190. [Online]. Available: http://dx.doi.org/10.1007/978-3-54079996-2 10
[55] S. Patidar, D. Rane, and P. Jain, A survey paper on cloud computing,

in Advanced Computing Communication Technologies (ACCT), 2012
Second International Conference on, jan. 2012, pp. 394 398.
[Online]. Available: http://dx.doi.org/10.1109/ACCT.2012.15
[56] M. Zhou, R. Zhang, D. Zeng, and W. Qian, Services in the cloud
computing era: A survey, in Universal Communication Symposium
(IUCS), 2010 4th International, oct. 2010, pp. 40 46. [Online].
Available: http://dx.doi.org/10.1109/IUCS.2010.5666772
[57] P. Banerjee, R. Friedrich, C. Bash, P. Goldsack, B. Huberman,
J. Manley, C. Patel, P. Ranganathan, and A. Veitch, Everything
as a service: Powering the new information economy, Computer,
vol. 44, no. 3, pp. 36 43, march 2011. [Online]. Available:
http://dx.doi.org/10.1109/MC.2011.67
[58] V. Issarny, M. Caporuscio, and N. Georgantas, A perspective on the
future of middleware-based software engineering, in 2007 Future
of Software Engineering, ser. FOSE 07. Washington, DC, USA:
IEEE Computer Society, 2007, pp. 244258. [Online]. Available:
http://dx.doi.org/10.1109/FOSE.2007.2
[59] M. Nagy, A. Katasonov, O. Khriyenko, S. Nikitin, M. Szydlowski,
and V. Terziyan, Challenges of middleware for the internet
of things, University of Jyvaskyla, Tech. Rep., 2009,
http://cdn.intechopen.com/pdfs/8786/InTech-Challenges of
middleware for the internet of things.pdf [Accessed on: 201112-20].
[60] OpenIoT Consortium, Open source solution for the internet of things
into the cloud, January 2012, http://www.openiot.eu [Accessed on:
2012-04-08].
[61] A. Badii, M. Crouch, and C. Lallah, A context-awareness framework
for intelligent networked embedded systems, in Advances in HumanOriented and Personalized Mechanisms, Technologies and Services
(CENTRIC), 2010 Third International Conference on, aug. 2010, pp.
105 110. [Online]. Available: http://dx.doi.org/10.1109/CENTRIC.
2010.29
[62] M. Simonov, Ismb middleware for iot (rfid), Presentation
Slides, March 2010, http://ec.europa.eu/information-society/activities/
foi/events/fippp/docs/mikhail-simonov.pdf [Accessed on: 2012-04-04].
[63] N. Kefalakis, N. Leontiadis, J. Soldatos, and D. Donsez, Middleware
building blocks for architecting rfid systems, in Mobile Lightweight
Wireless Systems, ser. Lecture Notes of the Institute for Computer
Sciences, Social Informatics and Telecommunications Engineering,
F. Granelli, C. Skianis, P. Chatzimisios, Y. Xiao, S. Redana,
O. Akan, P. Bellavista, J. Cao, F. Dressler, D. Ferrari, M. Gerla,
H. Kobayashi, S. Palazzo, S. Sahni, X. S. Shen, M. Stan,
J. Xiaohua, A. Zomaya, and G. Coulson, Eds. Springer Berlin
Heidelberg, 2009, vol. 13, pp. 325336. [Online]. Available:
http://dx.doi.org/10.1007/978-3-642-03819-8 31
[64] A. Katasonov, O. Kaykova, O. Khriyenko, S. Nikitin, and V. Y.
Terziyan, Smart semantic middleware for the internet of things.
in ICINCO-ICSO08, 2008, pp. 169178. [Online]. Available:
http://www.mit.jyu.fi/ai/papers/ICINCO-2008.pdf
[65] M. Caporuscio, P.-G. Raverdy, and V. Issarny, ubisoap: A serviceoriented middleware for ubiquitous networking, IEEE Trans. Services
Computing, vol. 5, no. 1, pp. 86 98, jan.-march 2012. [Online].
Available: http://dx.doi.org/10.1109/TSC.2010.60
[66] V. Terziyan, O. Kaykova, and D. Zhovtobryukh, Ubiroad: Semantic
middleware for context-aware smart road environments, in Internet
and Web Applications and Services (ICIW), 2010 Fifth International
Conference on, may 2010, pp. 295 302. [Online]. Available:
http://dx.doi.org/10.1109/ICIW.2010.50
[67] A. Salehi, Design and implementation of an efficient data stream
processing system. Ph.D. dissertation, Ecole Polytechnique Federale
de Lausanne (EPFL), 2010, http://biblion.epfl.ch/EPFL/theses/2010/
4611/EPFL TH4611.pdf [Accessed on: 2011-10-05].
[68] M. Albano, A. Brogi, R. Popescu, M. Diaz, and J. A. Dianes, Towards
secure middleware for embedded peer-to-peer systems: Objectives
and requirements, in RSPSI 07: Workshop on Requirements and
Solutions for Pervasive Software Infrastructures, 2007, pp. 1
6. [Online]. Available: http://citeseerx.ist.psu.edu/viewdoc/download?
doi=10.1.1.90.5982&rep=rep1&type=pdf
[69] A. Cannata, M. Gerosa, and M. Taisch, Socrades: A framework
for developing intelligent systems in manufacturing, in Industrial
Engineering and Engineering Management, 2008. IEEM 2008. IEEE
International Conference on, dec. 2008, pp. 1904 1908. [Online].
Available: http://dx.doi.org/10.1109/IEEM.2008.4738203
[70] H. Bohn, A. Bobek, and F. Golatowski, Sirena - service infrastructure
for real-time embedded networked devices: A service oriented
framework for different domains, in Networking, International
Conference on Systems and International Conference on Mobile
[71]
[72]
[73]
[74]
[75]
[76]
[77]
[78]
[79]
[80]
[81]
[82]
[83]
[84]
[85]
[86]
[87]
[88]
[89]
449
Communications and Learning Technologies, 2006. ICN/ICONS/MCL

2006. International Conference on, april 2006, p. 43. [Online].
Available: http://dx.doi.org/10.1109/ICNICONSMCL.2006.196
D. Giusto, A. Iera, G. Morabito, L. Atzori, A. Puliafito, A. Cucinotta,
A. L. Minnolo, and A. Zaia, Making the internet of things a reality:
The wherex solution, in The Internet of Things. Springer New York,
2010, pp. 99108. [Online]. Available: http://dx.doi.org/10.1007/9781-4419-1674-7 10
A. K. Dey, G. D. Abowd, and D. Salber, A conceptual
framework and a toolkit for supporting the rapid prototyping of
context-aware applications, Hum.-Comput. Interact., vol. 16, pp.
97166, December 2001. [Online]. Available: http://dx.doi.org/10.
1207/S15327051HCI16234 02
P. J. Brown, The stick-e document: a framework for creating
context-aware applications, Electronic Publishing, vol. 8, no. 2 &
3, pp. 259272, 1996. [Online]. Available: http://citeseerx.ist.psu.edu/
viewdoc/download?doi=10.1.1.8.7472&rep=rep1&type=pdf
D. Franklin and J. Flachsbart, All gadget and no representation
makes jack a dull environment, Proc. AAAI 1998 Spring Symposium
on Intelligent Environments, vol. Technical Report SS-98-02, pp.
155160, 1998. [Online]. Available: http://infolab.northwestern.edu/
media/papers/paper10072.pdf
T. Rodden, K. Chervest, N. Davies, and A. Dix, Exploiting context
in hci design for mobile systems, in in Workshop on Human
Computer Interaction with Mobile Devices, 1998. [Online]. Available:
http://eprints.lancs.ac.uk/11619/
R. Hull, P. Neaves, and J. Bedford-Roberts, Towards situated
computing, in Wearable Computers, 1997. Digest of Papers., First
International Symposium on, oct 1997, pp. 146 153. [Online].
Available: http://dx.doi.org/10.1109/ISWC.1997.629931
A. Ward, A. Jones, and A. Hopper, A new location technique for the
active office, IEEE Pers. Commun., vol. 4, no. 5, pp. 42 47, oct
1997. [Online]. Available: http://dx.doi.org/10.1109/98.626982
G. D. Abowd and E. D. Mynatt, Charting past, present, and
future research in ubiquitous computing, ACM Trans. Comput.-Hum.
Interact., vol. 7, pp. 2958, March 2000. [Online]. Available:
http://doi.acm.org/10.1145/344949.344988
B. Schilit, N. Adams, and R. Want, Context-aware computing
applications, in Mobile Computing Systems and Applications, 1994.
Proceedings., Workshop on, dec 1994, pp. 85 90. [Online]. Available:
http://dx.doi.org/10.1109/MCSA.1994.512740
J. Pascoe, Adding generic contextual capabilities to wearable
computers, in Wearable Computers, 1998. Digest of Papers. Second
International Symposium on, oct 1998, pp. 92 99. [Online]. Available:
http://dx.doi.org/10.1109/ISWC.1998.729534
Dictionary.com LLC, Thesaurus.com, 1995, http://thesaurus.com/
[accessed on: 2012-05-15].
Foldoc.org, Free on-line dictionary of computing, 2010, http://foldoc.
org/context [Accessed on: 2012-05-21].
princeton.edu, Wordnet: Lexical database for the english language,
2005, http://wordnetweb.princeton.edu/ [Accessed on: 2012-05-10].
Pearson Education Limited, Longman dictionary of contemporary
english advanced learners dictionary. 2012, http://www.ldoceonline.
com/dictionary/context [Accessed on: 2012-05-10].
L. Sanchez, J. Lanza, R. Olsen, M. Bauer, and M. Girod-Genet,
A generic context management framework for personal networking
environments, in Mobile and Ubiquitous Systems - Workshops, 2006.
3rd Annual International Conference on, july 2006, pp. 1 8. [Online].
Available: http://dx.doi.org/10.1109/MOBIQW.2006.361743
S. Ahn and D. Kim, Proactive context-aware sensor networks,
in Wireless Sensor Networks, ser. Lecture Notes in Computer
Science, K. Rmer, H. Karl, and F. Mattern, Eds. Springer Berlin
/ Heidelberg, 2006, vol. 3868, pp. 3853. [Online]. Available:
http://dx.doi.org/10.1007/11669463 6
N. S. Ryan, J. Pascoe, and D. R. Morse, Enhanced reality
fieldwork: the context-aware archaeological assistant, in Computer
Applications in Archaeology 1997, ser. British Archaeological
Reports, V. Gaffney, M. van Leusen, and S. Exxon, Eds.
Oxford: Tempus Reparatum, October 1998. [Online]. Available:
http://www.cs.kent.ac.uk/pubs/1998/616
P. Hu, J. Indulska, and R. Robinson, An autonomic context
management system for pervasive computing, in Pervasive Computing
and Communications, 2008. PerCom 2008. Sixth Annual IEEE
International Conference on, march 2008, pp. 213 223. [Online].
Available: http://dx.doi.org/10.1109/PERCOM.2008.56
K. Henricksen, A framework for context-aware pervasive computing
applications, Computer Science, School of Information Technology
and Electrical Engineering, The University of Queensland, Septem-
450
[90]
[91]
[92]
[93]
[94]
[95]
[96]
[97]
[98]
[99]
[100]
[101]
[102]
[103]
[104]
[105]
ber 2003, http://henricksen.id.au/publications/phd-thesis.pdf [Accessed

on:2012-01-05].
A. Moses, Lg smart fridge tells you what to buy, cook
and eat, The Sydney Morning Herald, January 2012,
http://www.smh.com.au/digital-life/hometech/lg-smart-fridge-tellsyou-what-to-buy-cook-and-eat-20120110-1ps9z.html [Accessed on:
2012-04-04].
M. Raskino, J. Fenn, and A. Linden, Extracting value from
the massively connected world of 2015, Gartner Research, Tech.
Rep., April 2005, http://www.gartner.com/resources/125900/125949/
extracting valu.pdf [Accessed on: 2011-08-20].
P. Prekop and M. Burnett, Activities, context and ubiquitous
computing, Special Issue on Ubiquitous Computing Computer
Communications, vol. 26, no. 11, p. 11681176, 2003. [Online].
Available: http://arxiv.org/ftp/cs/papers/0209/0209021.pdf
R. M. Gustavsen, Condor - an application framework for
mobility-based context-aware applications, in Proc. Workshop on
Concepts and Models for Ubiquitous Computing, Goeteborg, Sweden,
2002. [Online]. Available: http://www.alandix.com/academic/conf/
ubicomp2002-models/pdf/Gustavsen-goteborg%20sept-02.pdf
T. Hofer, W. Schwinger, M. Pichler, G. Leonhartsberger, J. Altmann,
and W. Retschitzegger, Context-awareness on mobile devices - the
hydrogen approach, in Proc. 36th Annual Hawaii International
Conference on System Sciences, ser. HICSS 03, 2003, p. 292302.
[Online]. Available: http://dx.doi.org/10.1109/HICSS.2003.1174831
A. van Bunningen, L. Feng, and P. Apers, Context for ubiquitous
data management, in Ubiquitous Data Management, 2005. UDM
2005. International Workshop on, april 2005, pp. 17 24. [Online].
Available: http://dx.doi.org/10.1109/UDM.2005.7
Z. Miao and B. Yuan, Spontaneous sensor networks for context-aware
computing, in Wireless, Mobile and Multimedia Networks, 2006 IET
International Conference on, nov. 2006, pp. 1 4. [Online]. Available:
http://ieeexplore.ieee.org/xpls/abs all.jsp?arnumber=5195749
D. Guan, W. Yuan, S. Lee, and Y.-K. Lee, Context selection
and reasoning in ubiquitous computing, in Intelligent Pervasive
Computing, 2007. IPC. The 2007 International Conference on, oct.
IPC.2007.102
S. K. Chong, I. McCauley, S. W. Loke, and S. Krishnaswamy,
Context-aware
sensors
and
data muling,
in
Context
awareness for self-managing systems (devices, applications and
networks) proceeding. Berlin : VDE-Verlag, 2007, pp. 103
117. [Online]. Available: http://arrow.latrobe.edu.au:8080/vital/access/
HandleResolver/1959.9/122313
G. Jun-zhong, Context aware computing, Journal of East China
Normal University (Natural Science), vol. 5, pp. 120, 2009.
[Online]. Available: http://en.cnki.com.cn/Article en/CJFDTOTALHDSZ200905002.htm
L. Mei and S. Easterbrook, Capturing and modeling human cognition
for context-aware software, in International conference for research
on computational models and computation-based theories of human
behavior, 2009. [Online]. Available: http://sideshow.psyc.bbk.ac.uk/
rcooper/iccm2009/proceedings/papers/0090/paper0090.pdf
S. Rizou, K. Haussermann, F. Durr, N. Cipriani, and K. Rothermel,
A system for distributed context reasoning, in Autonomic and
Autonomous Systems (ICAS), 2010 Sixth International Conference on,
march 2010, pp. 84 89. [Online]. Available: http://dx.doi.org/10.
1109/ICAS.2010.21
W. Liu, X. Li, and D. Huang, A survey on context awareness,
in Computer Science and Service System (CSSS), 2011 International
Conference on, june 2011, pp. 144 147. [Online]. Available:
http://dx.doi.org/10.1109/CSSS.2011.5972040
S. Yanwei, Z. Guangzhou, and P. Haitao, Research on the context
model of intelligent interaction system in the internet of things,
in IT in Medicine and Education (ITME), 2011 International
Symposium on, vol. 2, dec. 2011, pp. 379 382. [Online]. Available:
http://dx.doi.org/10.1109/ITiME.2011.6132129
L. Barkhuus, L. Barkhuus, and A. Dey, Is context-aware computing
taking control away from the user? three levels of interactivity
examined, in In Proc. Ubicomp 2003. Springer, 2003, pp. 149
156. [Online]. Available: http://www.itu.dk/people/barkhuus/barkhuus
ubicomp.pdf
D. Martin, C. Lamsfus, and A. Alzua, Automatic context data
life cycle management framework, in Pervasive Computing and
Applications (ICPCA), 2010 5th International Conference on, dec.
ICPCA.2010.5704122
[106] F. Ramparany, R. Poortinga, M. Stikic, J. Schmalenstroer, and

T. Prante, An open context information management infrastructure
the ist-amigo project, in Intelligent Environments, 2007. IE 07. 3rd
IET International Conference on, sept. 2007, pp. 398 403. [Online].
Available: http://dx.doi.org/10.1049/cp:20070398
[107] A. Bernardos, P. Tarrio, and J. Casar, A data fusion framework for
context-aware mobile services, in Multisensor Fusion and Integration
for Intelligent Systems, 2008. MFI 2008. IEEE International
Conference on, aug. 2008, pp. 606 613. [Online]. Available:
http://dx.doi.org/10.1109/MFI.2008.4648011
[108] B. Y. Lim and A. K. Dey, Toolkit to support intelligibility
in context-aware applications, in Proc. 12th ACM international
conference on Ubiquitous computing, ser. Ubicomp 10. New
http://doi.acm.org/10.1145/1864349.1864353
[109] G. Hynes, V. Reynolds, and M. Hauswirth, A context lifecycle
for web-based context management services, in Smart Sensing
and Context, ser. Lecture Notes in Computer Science, P. Barnaghi,
K. Moessner, M. Presser, and S. Meissner, Eds. Springer Berlin
/ Heidelberg, 2009, vol. 5741, pp. 5165. [Online]. Available:
http://dx.doi.org/10.1007/978-3-642-04471-7 5
[110] M. Peterson and E. Pierre, Snias vision for information life cycle
management (ilm), in Storage Networking World. Computer World,
2004.
[111] AIIM, What is enterprise content management (ecm)? February
2009,
http://www.aiim.org/What-is-ECM-Enterprise-ContentManagement.aspx [Accessed on: 2012-06-20].
[112] E. Hayden, Data lifecycle management model shows risks and integrated data flow, in Information Security Magazine, July 2008.
[113] A. N. Shulsky and G. J. Schmitt, Silent Warfare: Understanding the
World of Intelligence, 3rd ed. Potomac Books Inc, May 2002.
[114] J. R. Boyd, A discourse on winning and losing. Unpublished set
of briefing slides available at Air University Library, Maxwell AFB,
Alabama, 1987, http://www.ausairpower.net/JRB/intro.pdf [Accessed:
2011-12-18].
[115] M. Chantzara and M. Anagnostou, Evaluation and selection of
context information, in In: Second International Workshop on
Modeling and Retrieval of Context, Edinburgh, 2005. [Online].
Available: http://ceur-ws.org/Vol-146/paper7.pdf
[116] A. Ferscha, S. Vogl, and W. Beer, Context sensing, aggregation,
representation and exploitation in wireless networks, Scalable
Computing: Practice and Experience, vol. 6, no. 2, p. 7181, 2005.
[Online]. Available: http://www.scpe.org/index.php/scpe/article/view/
327/17
[117] K. Wrona and L. Gomez, Context-aware security and secure
context-awareness in ubiquitous computing environments, in XXI
Autumn Meeting of Polish Information Processing Society, 2005.
[Online]. Available: http://proceedings2005.imcsit.org/docs/75.pdf
[118] S. Pietschmann, A. Mitschick, R. Winkler, and K. Meissner,
Croco: Ontology-based, cross-application context management, in
Semantic Media Adaptation and Personalization, 2008. SMAP 08.
Third International Workshop on, dec. 2008, pp. 88 93. [Online].
Available: http://dx.doi.org/10.1109/SMAP.2008.10
[119] H. Chen, T. Finin, A. Joshi, L. Kagal, F. Perich, and D. Chakraborty,
Intelligent agents meet the semantic web in smart spaces, IEE
Internet Computing, vol. 8, no. 6, pp. 69 79, nov.-dec. 2004.
[Online]. Available: http://dx.doi.org/10.1109/MIC.2004.66
[120] J. Indulska and P. Sutton, Location management in pervasive
systems, in Proc. Australasian information security workshop
conference on ACSW frontiers 2003 - Volume 21, ser. ACSW Frontiers
03. Darlinghurst, Australia, Australia: Australian Computer Society,
Inc., 2003, pp. 143151. [Online]. Available: http://dl.acm.org/citation.
cfm?id=827987.828003
[121] A. Schmidt and K. van Laerhoven, How to build smart appliances?
IEEE Pers. Commun., vol. 8, no. 4, pp. 66 71, aug. 2001. [Online].
Available: http://dx.doi.org/10.1109/98.944006
[122] K. Balavalad, S. Manvi, and A. Sutagundar, Context aware computing
in wireless sensor networks, in Advances in Recent Technologies in
Communication and Computing, 2009. ARTCom 09. International
Conference on, oct. 2009, pp. 514 516. [Online]. Available:
http://dx.doi.org/10.1109/ARTCom.2009.85
[123] G. Castelli, M. Mamei, A. Rosi, and F. Zambonelli, Extracting
high-level information from location data: the w4 diary example,
Mob. Netw. Appl., vol. 14, no. 1, pp. 107119, Feb. 2009. [Online].
Available: http://dx.doi.org/10.1007/s11036-008-0104-y
[124] H. Chang, S. Shin, and C. Chung, Context life cycle management
scheme in ubiquitous computing environments, in Mobile Data
[125]
[126]
[127]
[128]
[129]
[130]
[131]
[132]
[133]
[134]
[135]
[136]
[137]
[138]
[139]
[140]
[141]
Management, 2007 International Conference on, may 2007, pp. 315

319. [Online]. Available: http://dx.doi.org/10.1109/MDM.2007.66
W3C Ubiquitos Web Domain, Composite capabilities/preference profiles: Structure and vocabularies 2.0, June 2007, http://www.w3.org/
Mobile/CCPP/ [Accessed on: 2012-05-26].
M. Knappmeyer, S. L. Kiani, C. Fra, B. Moltchanov, and N. Baker,
Contextml: A light-weight context representation and context
management schema, in Wireless Pervasive Computing (ISWPC),
2010 5th IEEE International Symposium on, may 2010, pp. 367 372.
[Online]. Available: http://dx.doi.org/10.1109/ISWPC.2010.5483753
uml.org, Unified modeling language (uml), 2012, http://www.uml.
org/ [Addressed on: 2012-05-29].
ormfoundation.org,
The
orm
foundation,
1989,
http:
//www.ormfoundation.org [Addressed on: 2012-05-29].
J. Han, E. Haihong, G. Le, and J. Du, Survey on nosql
database, in Pervasive Computing and Applications (ICPCA), 2011
6th International Conference on, oct. 2011, pp. 363 366. [Online].
Available: http://dx.doi.org/10.1109/ICPCA.2011.6106531
D. Allemang and J. Hendler, Semantic Web for the Working
Ontologist, Second Edition: Effective Modeling in RDFS and
OWL, 2nd ed. Morgan Kaufmann, 2011. [Online]. Available:
http://mkp.com/news/semantic-web-for-the-working-ontologist-2ndedition-effective-modeling-in-rdfs-and-owl-by-dean-allemang-jameshendler
L.
Yu,
A
Developers
Guide
to
the
Semantic
Web,
1st
ed.
Springer,
2011.
[Online].
Available: http://www.springer.com/computer/database+management+%26+
information+retrieval/book/978-3-642-15969-5
P. Hitzler, M. Krtzsch, and S. Rudolph, Foundations of Semantic
Web Technologies.
Chapman & Hall/CRC, 2009. [Online].
Available: http://www.semantic-web-book.org/page/Foundations of
Semantic Web Technologies
M. Botts and A. Robin, Opengis sensor model language (sensorml)
implementation specification, Open Geospatial Consortium Inc, Tech.
Rep., 2007, https://portal.opengeospatial.org/modules/admin/license
agreement.php?suppressHeaders=0&access license id=3&target=http:
//portal.opengeospatial.org/files/%3fartifact id=12606 [Accessed on:
2011-12-15].
B. Khoo, Rfid- from tracking to the internet of things: A review
of developments, in Proc. 2010 IEEE/ACM Intl Conference on
Green Computing and Communications & Intl Conference on
Cyber, Physical and Social Computing, ser. GREENCOM-CPSCOM
10. Washington, DC, USA: IEEE Computer Society, 2010, pp.
533538. [Online]. Available: http://dx.doi.org/10.1109/GreenComCPSCom.2010.22
X. H. Wang, T. Gu, D. Q. Zhang, and H. K. Pung, Ontology based
context modeling and reasoning using owl, in Pervasive Computing
and Communications Workshops, 2004. Proc. Second IEEE Annual
Conference on, march 2004, pp. 18 22. [Online]. Available:
http://dx.doi.org/10.1109/PERCOMW.2004.1276898
N. F. Noy and D. L. McGuinness, Ontology development 101: A
guide to creating your first ontology, Stanford University, Stanford,
CA, Tech. Rep., March 2001, http://protege.stanford.edu/publications/
ontology development/ontology101.pdf [Accessed on: 2011-12-15].
R. Studer, V. Benjamins, and D. Fensel, Knowledge engineering:
Principles and methods, Data & Knowledge Engineering,
vol. 25, no. 12, pp. 161 197, 1998. [Online]. Available:
http://dx.doi.org/10.1016/S0169-023X(97)00056-6
M. Uschold and M. Gruninger, Ontologies: Principles, methods
and applications, The Knowledge Engineering Review, vol. 11,
no. 02, pp. 93136, 1996. [Online]. Available: http://dx.doi.org/10.
1017/S0269888900007797
P. Korpipaa, J. Mantyjarvi, J. Kela, H. Keranen, and E.-J. Malm,
Managing context information in mobile devices, IEEE Pervasive
Computing, vol. 2, no. 3, pp. 4251, Jul. 2003. [Online]. Available:
P. Korpipaa and J. Mantyjarvi, An ontology for mobile device
sensor-based context awareness, in Proc. 4th international and
interdisciplinary conference on Modeling and using context, ser.
CONTEXT03. Berlin, Heidelberg: Springer-Verlag, 2003, pp. 451
458. [Online]. Available: http://dl.acm.org/citation.cfm?id=1763142.
1763181
K.-E. Ko and K.-B. Sim, Development of context aware system based
on bayesian network driven context reasoning method and ontology
context modeling, in Control, Automation and Systems, 2008. ICCAS
2008. International Conference on, oct. 2008, pp. 2309 2313.
[Online]. Available: http://dx.doi.org/10.1109/ICCAS.2008.4694191
451
[142] R. de Freitas Bulcao Neto and M. da Graca Campos Pimentel,

Toward a domain-independent semantic model for context-aware
computing, in Web Congress, 2005. LA-WEB 2005. Third Latin
American, oct.-2 nov. 2005, p. 10 pp. [Online]. Available:
http://dx.doi.org/10.1109/LAWEB.2005.43
[143] M. Compton, C. Henson, H. Neuhaus, L. Lefort, and A. Sheth, A
survey of the semantic specification of sensors, in 2nd International
Workshop on Semantic Sensor Networks, at 8th International Semantic
Web Conference,, Oct. 2009. [Online]. Available: http://sunsite.
informatik.rwth-aachen.de/Publications/CEUR-WS/Vol-522/p6.pdf
[144] w3.org, Resource description framework (rdf), 2004, http://www.w3.
org/RDF/ [accesed on: 2012-01-21].
, Rdf vocabulary description language 1.0: Rdf schema (rdfs),
[145]
2004, http://www.w3.org/2001/sw/wiki/RDFS [accesed on: 2012-0121].
[146] W3.org, Web ontology language (owl), Webpage, October 2007,
www.w3.org/2004/OWL/ [Accessed: 2011-12-18].
[147] A. Bikakis, T. Patkos, G. Antoniou, , and D. Plexousaki, A survey
of semantics-based approaches for context reasoning in ambient
intelligence, in Ambient Intelligence 2007 Workshops, M. M, F. A, and
A. E, Eds., vol. 11. SPRINGER-VERLAG BERLIN, 2008. [Online].
Available: http://www.csd.uoc.gr/ bikakis/pubs/survey-ami07.pdf
[148] P. Nurmi and P. Floree. (2004) Reasoning in context-aware systems. Position Paper. Department of Computer Science, University of
Helsinki.
[149] D. Hall and J. Llinas, An introduction to multisensor data fusion,
Proc. IEEE, vol. 85, no. 1, pp. 6 23, jan 1997. [Online]. Available:
http://dx.doi.org/10.1109/5.554205
[150] N. Lane, E. Miluzzo, H. Lu, D. Peebles, T. Choudhury, and
A. Campbell, A survey of mobile phone sensing, IEEE Commun.
Mag., vol. 48, no. 9, pp. 140 150, sept. 2010. [Online]. Available:
http://dx.doi.org/10.1109/MCOM.2010.5560598
[151] D. Riboni and C. Bettini, Context-aware activity recognition through
a combination of ontological and statistical reasoning, in Proceedings
of the 6th International Conference on Ubiquitous Intelligence and
Computing, ser. UIC 09. Berlin, Heidelberg: Springer-Verlag, 2009,
pp. 3953. [Online]. Available: http://dx.doi.org/10.1007/978-3-64202830-4 5
[152] S.-H. Huang, T.-T. Wu, H.-C. Chu, and G.-J. Hwang, A decision
tree approach to conducting dynamic assessment in a contextaware ubiquitous learning environment, in Wireless, Mobile, and
Ubiquitous Technology in Education, 2008. WMUTE 2008. Fifth
IEEE International Conference on, march 2008, pp. 89 94. [Online].
Available: http://dx.doi.org/10.1109/WMUTE.2008.10
[153] H.-S. Park, K. Oh, and S.-B. Cho, Bayesian network-based
high-level context recognition for mobile context sharing in
cyber-physical system, International Journal of Distributed Sensor
Networks, vol. 2011, p. 10, 2011. [Online]. Available: http:
//downloads.hindawi.com/journals/ijdsn/2011/650387.pdf
[154] B. Korel and S. Koo, A survey on context-aware sensing for
body sensor networks, Scientific Research Publishing: Wireless
Sensor Network, vol. 2 (8), pp. 571583, 2010. [Online]. Available:
http://www.scirp.org/journal/PaperDownload.aspx?paperID=2345
[155] C. Doukas, I. Maglogiannis, P. Tragas, D. Liapis, and G. Yovanof,
Patient fall detection using support vector machines, in Artificial
Intelligence and Innovations 2007: from Theory to Applications,
ser. IFIP International Federation for Information Processing,
C. Boukis, A. Pnevmatikakis, and L. Polymenakos, Eds. Springer
Boston, 2007, vol. 247, pp. 147156. [Online]. Available: http:
//dx.doi.org/10.1007/978-0-387-74161-1 16
[156] O. Brdiczka, J. Crowley, and P. Reignier, Learning situation
models in a smart home, IEEE Trans. Syst. Man Cybern. B,
Cybern, vol. 39, no. 1, pp. 56 63, feb. 2009. [Online]. Available:
http://dx.doi.org/10.1109/TSMCB.2008.923526
[157] T.-N. Lin and P.-C. Lin, Performance comparison of indoor
positioning techniques based on location fingerprinting in wireless
networks, in Wireless Networks, Communications and Mobile
Computing, 2005 International Conference on, vol. 2, june 2005,
pp. 1569 1574 vol.2. [Online]. Available: http://dx.doi.org/10.1109/
WIRLES.2005.1549647
[158] K. V. Laerhoven, Combining the self-organizing map and kmeans clustering for on-line classification of sensor data, in Proc.
International Conference on Artificial Neural Networks, ser. ICANN
01. London, UK, UK: Springer-Verlag, 2001, pp. 464469. [Online].
Available: http://dl.acm.org/citation.cfm?id=646258.683987
[159] R. Shtykh and Q. Jin, Capturing user contexts: Dynamic profiling for
information seeking tasks, in Systems and Networks Communications,
452
[160]
[161]
[162]
[163]
[164]
[165]
[166]
[167]
[168]
[169]
[170]
[171]
[172]
[173]
[174]
[175]
2008. ICSNC 08. 3rd International Conference on, oct. 2008, pp. 365
370. [Online]. Available: http://dx.doi.org/10.1109/ICSNC.2008.55
w3.org, Swrl: A semantic web rule language combining owl and
ruleml, May 2004, http://www.w3.org/Submission/SWRL/ [Accessed
on:2012-01-03].
X. Zhou, X. Tang, X. Yuan, and D. Chen, Spbca: Semantic patternbased context-aware middleware, in Parallel and Distributed Systems
(ICPADS), 2009 15th International Conference on, dec. 2009, pp. 891
895. [Online]. Available: http://dx.doi.org/10.1109/ICPADS.2009.146
C. Kessler, M. Raubal, and C. Wosniok, Semantic rules for contextaware geographical information retrieval, in Proc. 4th European
conference on Smart sensing and context, ser. EuroSSC09. Berlin,
Heidelberg: Springer-Verlag, 2009, pp. 7792. [Online]. Available:
http://dx.doi.org/10.1007/978-3-642-04471-7 7
C. Choi, I. Park, S. Hyun, D. Lee, and D. Sim, Mire: A
minimal rule engine for context-aware mobile devices, in Digital
Information Management, 2008. ICDIM 2008. Third International
Conference on, nov. 2008, pp. 172 177. [Online]. Available:
http://dx.doi.org/10.1109/ICDIM.2008.4746772
C. Barbero, P. D. Zovo, and B. Gobbi, A flexible context
aware reasoning approach for iot applications, in Proc. 2011
IEEE 12th International Conference on Mobile Data Management
- Volume 01, ser. MDM 11. Washington, DC, USA: IEEE
Computer Society, 2011, pp. 266275. [Online]. Available: http:
//dx.doi.org/10.1109/MDM.2011.55
K. Teymourian, O. Streibel, A. Paschke, R. Alnemr, and C. Meinel,
Towards semantic event-driven systems, in Proc. 3rd international
conference on New technologies, mobility and security, ser. NTMS09.
Piscataway, NJ, USA: IEEE Press, 2009, pp. 347352. [Online].
Available: http://dx.doi.org/10.1109/NTMS.2009.5384713
N. Konstantinou, E. Solidakis, S. Zoi, A. Zafeiropoulos,
P. Stathopoulos, and N. Mitrou, Priamos: a middleware
architecture for real-time semantic annotation of context features,
in Intelligent Environments, 2007. IE 07. 3rd IET International
Conference on, sept. 2007, pp. 96 103. [Online]. Available:
http://ieeexplore.ieee.org/xpls/abs all.jsp?arnumber=4449917
T. Terada, M. Tsukamoto, K. Hayakawa, T. Yoshihisa, Y. Kishino,
A. Kashitani, and S. Nishio, Ubiquitous chip: A rule-based i/o
control device for ubiquitous computing, in Pervasive Computing,
ser. Lecture Notes in Computer Science, A. Ferscha and F. Mattern,
Eds. Springer Berlin Heidelberg, 2004, vol. 3001, pp. 238253.
[Online]. Available: http://dx.doi.org/10.1007/978-3-540-24646-6 18
M. Roman, C. Hess, R. Cerqueira, A. Ranganathan, R. H. Campbell,
and K. Nahrstedt, A middleware infrastructure for active spaces,
IEEE Pervasive Computing, vol. 1, no. 4, pp. 7483, Oct. 2002.
[Online]. Available: http://dx.doi.org/10.1109/MPRV.2002.1158281
A. Ranganathan and R. H. Campbell, A middleware for contextaware agents in ubiquitous computing environments, in Proc.
ACM/IFIP/USENIX 2003 International Conference on Middleware,
ser. Middleware 03. New York, NY, USA: Springer-Verlag
New York, Inc., 2003, pp. 143161. [Online]. Available: http:
//dl.acm.org/citation.cfm?id=1515915.1515926
J. Mantyjarvi and T. Seppanen, Adapting applications in mobile
terminals using fuzzy context information, in Proc. 4th International
Symposium on Mobile Human-Computer Interaction, ser. Mobile HCI
02. London, UK, UK: Springer-Verlag, 2002, pp. 95107. [Online].
Available: http://dx.doi.org/10.1007/3-540-45756-9 9
A. Padovitz, S. W. Loke, and A. Zaslavsky, The ecora framework:
A hybrid architecture for context-oriented pervasive computing,
Pervasive Mob. Comput., vol. 4, no. 2, pp. 182215, Apr. 2008.
[Online]. Available: http://dx.doi.org/10.1016/j.pmcj.2007.10.002
D. Tsarkov, Fact++, Software, 2007, http://owl.man.ac.uk/
factplusplus/ [Accessed on: 2012-01-21].
Clark andParsia, Pellet: Owl 2 reasoner for java, Software, 2004,
http://clarkparsia.com/pellet/ [Accessed: 2011-12-18].
A. Zafeiropoulos, N. Konstantinou, S. Arkoulis, D.-E. Spanos,
and N. Mitrou, A semantic-based architecture for sensor data
fusion, in Mobile Ubiquitous Computing, Systems, Services and
Technologies, 2008. UBICOMM 08. The Second International
Conference on, 29 2008-oct. 4 2008, pp. 116 121. [Online].
Available: http://dx.doi.org/10.1109/UBICOMM.2008.67
A. Zafeiropoulos, D.-E. Spano, S. Arkoulis, N. Konstantinou, and
N. Mitrou, Data Management in the Semantic Web, ser. Distributed,
Cluster and Grid Computing - Yi Pan (Georgia State University),
Series Edito, H. Jin, Ed. NOVA Publishers, 2011. [Online].
Available: https://www.novapublishers.com/catalog/product info.php?
products id=20094
[176] Z. Song, A. Ca andrdenas, and R. Masuoka, Semantic middleware

for the internet of things, in Internet of Things (IOT), 2010,
29 2010-dec. 1 2010, pp. 1 8. [Online]. Available: http:
//dx.doi.org/10.1109/IOT.2010.5678448
[177] L. Peizhi and Z. Jian, A context-aware application infrastructure
with reasoning mechanism based on dempster-shafer evidence
theory, in Vehicular Technology Conference, 2008. VTC Spring
2008. IEEE, may 2008, pp. 2834 2838. [Online]. Available:
http://dx.doi.org/10.1109/VETECS.2008.618
[178] C. H. Lyu, M. S. Choi, Z. Y. Li, and H. Y. Youn, Reasoning
with imprecise context using improved dempster-shafer theory, in
Web Intelligence and Intelligent Agent Technology (WI-IAT), 2010
IEEE/WIC/ACM International Conference on, vol. 2, 31 2010-sept. 3
2010, pp. 475 478. [Online]. Available: http://dx.doi.org/10.1109/WIIAT.2010.190
[179] D. Zhang, J. Cao, J. Zhou, and M. Guo, Extended dempster-shafer
theory in context reasoning for ubiquitous computing environments, in
Computational Science and Engineering, 2009. CSE 09. International
Conference on, vol. 2, aug. 2009, pp. 205 212. [Online]. Available:
http://dx.doi.org/10.1109/CSE.2009.201
[180] P. Blunsom, Hidden markov models, University of Melbourne,
tutorial, August 2004, http://digital.cs.usu.edu/ cyan/CS7960/hmmtutorial.pdf [Accessed on: 2012-07-10].
[181] M. Krause, C. Linnhoff-Popien, and M. Strassberger, Concurrent
inference on high level context using alternative context construction
trees, in Autonomic and Autonomous Systems, 2007. ICAS07. Third
International Conference on, june 2007, p. 7. [Online]. Available:
http://dx.doi.org/10.1109/CONIELECOMP.2007.67
[182] T. Buchholz, M. Krause, C. Linnhoff-Popien, and M. Schiffers,
Coco: dynamic composition of context information, in Mobile and
Ubiquitous Systems: Networking and Services, 2004. MOBIQUITOUS
2004. The First Annual International Conference on, aug. 2004,
pp. 335 343. [Online]. Available: http://dx.doi.org/10.1109/MOBIQ.
2004.1331740
[183] W. Dargie, The role of probabilistic schemes in multisensor contextawareness, in Proc. Fifth IEEE International Conference on Pervasive
Computing and Communications Workshops, ser. PERCOMW 07.
Washington, DC, USA: IEEE Computer Society, 2007, pp. 2732.
[Online]. Available: http://dx.doi.org/10.1109/PERCOMW.2007.115
[184] G. Chen, M. Li, and D. Kotz, Data-centric middleware for
context-aware pervasive computing, Pervasive Mob. Comput., vol. 4,
no. 2, pp. 216253, Apr. 2008. [Online]. Available: http://dx.doi.org/
10.1016/j.pmcj.2007.10.001
[185] A. Devaraju, S. Hoh, and M. Hartley, A context gathering
framework for context-aware mobile solutions, in Proc. 4th
international conference on mobile technology, applications, and
systems and the 1st international symposium on Computer human
interaction in mobile technology, ser. Mobility 07. New York,
NY, USA: ACM, 2007, pp. 3946. [Online]. Available: http:
//doi.acm.org/10.1145/1378063.1378070
[186] A. K. Dey, G. D. Abowd, and D. Salber, A context-based infrastructure
for smart environments, Georgia Institute of Technology, Tech. Rep.,
1999, http://www.cc.gatech.edu/fce/contexttoolkit/pubs/MANSE99.pdf
[Accessed on: 2011-12-05].
[187] O. Corcho and R. Garcia-Castro, Five challenges for the semantic
sensor web, Semant. web, vol. 1, no. 1,2, pp. 121125, Apr.
2010. [Online]. Available: http://dl.acm.org/citation.cfm?id=2019445.
2019450
[188] Q. Huaifeng and Z. Xingshe, Context aware sensornet, in
Proceedings of the 3rd international workshop on Middleware
for pervasive and ad-hoc computing, ser. MPAC 05. New
http://doi.acm.org/10.1145/1101480.1101489
[189] J. Filho and N. Agoulmine, A quality-aware approach for
resolving context conflicts in context-aware systems, in Embedded
and Ubiquitous Computing (EUC), 2011 IFIP 9th International
Conference on, oct. 2011, pp. 229 236. [Online]. Available:
http://dx.doi.org/10.1109/EUC.2011.9
[190] E. F. Nakamura, A. A. F. Loureiro, and A. C. Frery, Information fusion
for wireless sensor networks: Methods, models, and classifications,
ACM Comput. Surv., vol. 39, no. 3, pp. 9/19/55, Sep. 2007. [Online].
[191] D. Garlan, D. Siewiorek, A. Smailagic, and P. Steenkiste, Project
aura: Toward distraction-free pervasive computing, IEEE Pervasive
Computing, vol. 1, no. 2, pp. 2231, Apr. 2002. [Online]. Available:
[192] T. Strang, C. Linnhoff-Popien, and K. Frank, Cool: A context ontology
language to enable contextual interoperability, Ifip International
[193]
[194]
[195]
[196]
[197]
[198]
[199]
[200]
[201]
[202]
[203]
[204]
[205]
[206]
[207]
Federation For Information Processing, vol. 2893, pp. 236

247, 2003. [Online]. Available: http://www.springerlink.com/index/
DJNHU2GVPV7CQLTV.pdf
L. Capra, W. Emmerich, and C. Mascolo, Carisma: context-aware
reflective middleware system for mobile applications, IEEE Trans.
Softw. Eng., vol. 29, no. 10, pp. 929 945, oct. 2003. [Online].
Available: http://dx.doi.org/10.1109/TSE.2003.1237173
T. Gu, H. K. Pung, and D. Q. Zhang, A service-oriented
middleware for building context-aware services, J. Netw. Comput.
Appl., vol. 28, no. 1, pp. 118, Jan. 2005. [Online]. Available:
http://dx.doi.org/10.1016/j.jnca.2004.06.002
D. H. Wilson, A. C. Long, and C. Atkeson, A contextaware recognition survey for data collection using ubiquitous
sensors in the home, in CHI 05 extended abstracts on Human
factors in computing systems, ser. CHI EA 05. New York,
NY, USA: ACM, 2005, pp. 18651868. [Online]. Available:
http://doi.acm.org/10.1145/1056808.1057042
C.-H. Hou, H.-C. Hsiao, C.-T. King, and C.-N. Lu, Context
discovery in sensor networks, in Information Technology: Research
and Education, 2005. ITRE 2005. 3rd International Conference on,
june 2005, pp. 2 6. [Online]. Available: http://dx.doi.org/10.1109/
ITRE.2005.1503053
C. Jacob, D. Linner, S. Steglich, and I. Radusch, Bio-inspired
context gathering in loosely coupled computing environments,
in Bio-Inspired Models of Network, Information and Computing
Systems, 2006. 1st, dec. 2006, pp. 1 6. [Online]. Available:
http://dx.doi.org/10.1109/BIMNICS.2006.361803
D. Ejigu, M. Scuturici, and L. Brunie, Semantic approach to context
management and reasoning in ubiquitous context-aware systems, in
Digital Information Management, 2007. ICDIM 07. 2nd International
Conference on, vol. 1, oct. 2007, pp. 500 505. [Online]. Available:
http://dx.doi.org/10.1109/ICDIM.2007.4444272
R. de Rocha and M. Endler, Middleware: Context management
in heterogeneous, evolving ubiquitous environments, Distributed
Systems Online, IEEE, vol. 7, no. 4, p. 1, april 2006. [Online].
Available: http://dx.doi.org/10.1109/MDSO.2006.28
S.-H. Baek, E.-C. Choi, J.-D. Huh, and K.-R. Park, Sensor
information management mechanism for context-aware service
in ubiquitous home, IEEE Trans. Consum. Electron., vol. 53,
no. 4, pp. 1393 1400, nov. 2007. [Online]. Available: http:
//dx.doi.org/10.1109/TCE.2007.4429229
D. Conan, R. Rouvoy, and L. Seinturier, Scalable processing
of context information with cosmos, in Proc. 7th IFIP WG 6.1
international conference on Distributed applications and interoperable
systems, ser. DAIS07. Berlin, Heidelberg: Springer-Verlag, 2007,
pp. 210224. [Online]. Available: http://hal.inria.fr/docs/00/15/50/45/
PDF/article.pdf
J. Herbert, J. ODonoghue, and X. Chen, A context-sensitive
rule-based architecture for a smart building environment, in Future
Generation Communication and Networking, 2008. FGCN 08. Second
International Conference on, vol. 2, dec. 2008, pp. 437 440. [Online].
Available: http://dx.doi.org/10.1109/FGCN.2008.169
W. Xue, H. Pung, W. Ng, and T. Gu, Data management for
context-aware computing, in Embedded and Ubiquitous Computing,
2008. EUC 08. IEEE/IFIP International Conference on, vol. 1, dec.
EUC.2008.27
E. Kim and J. Choi, A context management system for supporting
context-aware applications, in Embedded and Ubiquitous Computing,
2008. EUC 08. IEEE/IFIP International Conference on, vol. 2, dec.
EUC.2008.168
Z. Li, X. Zhou, H. Qing, and S. Li, Model and implementation
of context-aware sensor networks, in Information Science and
Engineering, 2008. ISISE 08. International Symposium on, vol. 2,
dec. 2008, pp. 16 19. [Online]. Available: http://dx.doi.org/10.1109/
ISISE.2008.202
S. Gashti, G. Pujolle, and J. Rotrou, An upnp-based context-aware
framework for ubiquitous mesh home networks, in Personal, Indoor
and Mobile Radio Communications, 2009 IEEE 20th International
Symposium on, sept. 2009, pp. 400 404. [Online]. Available:
http://dx.doi.org/10.1109/PIMRC.2009.5449966
E. S. Reetz, R. Tonjes, and N. Baker, Towards global smart spaces:
Merge wireless sensor networks into context-aware systems, in
Wireless Pervasive Computing (ISWPC), 2010 5th IEEE International
Symposium on, may 2010, pp. 337 342. [Online]. Available:
http://dx.doi.org/10.1109/ISWPC.2010.5483728
453
[208] Y. Zhan, S. Wang, Z. Zhao, C. Chen, and J. Ma, A mobile

device oriented framework for context information management, in
Information, Computing and Telecommunication, 2009. YC-ICT 09.
IEEE Youth Conference on, sept. 2009, pp. 150 153. [Online].
Available: http://dx.doi.org/10.1109/YCICT.2009.5382404
[209] P. Pawar, H. Boros, F. Liu, G. Heijenk, and B.-J. van Beijnum,
Bridging context management systems in the ad hoc and mobile
environments, in Computers and Communications, 2009. ISCC 2009.
IEEE Symposium on, july 2009, pp. 882 888. [Online]. Available:
http://dx.doi.org/10.1109/ISCC.2009.5202323
[210] A. Corradi, M. Fanelli, and L. Foschini, Implementing a scalable
context-aware middleware, in Computers and Communications, 2009.
ISCC 2009. IEEE Symposium on, july 2009, pp. 868 874. [Online].
Available: http://dx.doi.org/10.1109/ISCC.2009.5202318
[211] O. Kwon, Y.-S. Song, J.-H. Kim, and K.-J. Li, Sconstream: A spatial
context stream processing system, in International Conference on
Computational Science and Its Applications, march 2010, pp. 165
170. [Online]. Available: http://dx.doi.org/10.1109/ICCSA.2010.48
[212] B. Guo, L. Sun, and D. Zhang, The architecture design of a
cross-domain context management system, in 8th IEEE International
Conference on Pervasive Computing and Communications Workshops,
29 2010-april 2 2010, pp. 499 504. [Online]. Available: http:
//dx.doi.org/10.1109/PERCOMW.2010.5470618
[213] W. Wibisono, A. Zaslavsky, and S. Ling, Comihoc: A middleware
framework for context management in manet environment, in
Advanced Information Networking and Applications (AINA), 2010
24th IEEE International Conference on, april 2010, pp. 620 627.
[Online]. Available: http://dx.doi.org/10.1109/AINA.2010.153
[214] A. Shaeib, P. Cappellari, and M. Roantree, A framework for real-time
context provision in ubiquitous sensing environments, in Computers
and Communications (ISCC), 2010 IEEE Symposium on, june 2010,
pp. 1083 1085. [Online]. Available: http://dx.doi.org/10.1109/ISCC.
2010.5546645
[215] F. Li, S. Sehic, and S. Dustdar, Copal: An adaptive approach to
context provisioning. in WiMob. IEEE, 2010, pp. 286293. [Online].
Available: http://dx.doi.org/10.1109/WIMOB.2010.5645051
[216] E. Badidi and I. Taleb, Towards a cloud-based framework for
context management, in Innovations in Information Technology (IIT),
2011 International Conference on, april 2011, pp. 35 40. [Online].
Available: http://dx.doi.org/10.1109/INNOVATIONS.2011.5893849
[217] A. K. Dey and A. Newberger, Support for context-aware intelligibility
and control, in Proc. 27th international conference on Human factors
in computing systems, ser. CHI 09. New York, NY, USA: ACM,
2009, pp. 859868. [Online]. Available: http://doi.acm.org/10.1145/
1518701.1518832
[218] jboss.org, Drools - the business logic integration platform, 2001, http:
//www.jboss.org/drools [Accessed on: 2012-04-23].
[219] A. Padovitz, S. Loke, and A. Zaslavsky, Towards a theory of context
spaces, in Pervasive Computing and Communications Workshops,
Proc. the Second IEEE Annual Conference on, march 2004, pp. 38
42. [Online]. Available: http://dx.doi.org/10.1109/PERCOMW.2004.
1276902
[220] T. Reineking, Java dempster shafer library, July 2010, https://
sourceforge.net/projects/jds/ [Accessed on: 2012-05-12].
[221] IEEE Instrumentation and Measurement Society, Ieee standard for
a smart transducer interface for sensors and actuators wireless
communication protocols and transducer electronic data sheet (teds)
formats, IEEE Std 1451.5-2007, pp. C1 236, 5 2007. [Online].
Available: http://dx.doi.org/10.1109/IEEESTD.2007.4346346
[222] P. Christen, Data Matching, ser. Data-Centric Systems and Applications. Springer, 2012.
[223] C. Perera, P. Jayaraman, A. Zaslavsky, P. Christen, and D. Georgakopoulos, Dynamic configuration of sensors using mobile sensor hub in internet of things paradigm, in IEEE 8th International
Conference on Intelligent Sensors, Sensor Networks, and Information
Processing (ISSNIP), Melbourne, Australia, April 2013, pp. 473478.
[224] C. Perera, A. Zaslavsky, P. Christen, A. Salehi, and D. Georgakopoulos, Connecting mobile things to global sensor network middleware
using system-generated wrappers, in International ACM Workshop
on Data Engineering for Wireless and Mobile Access 2012 (ACM
SIGMOD/PODS 2012-Workshop-MobiDE), Scottsdale, Arizona, USA,
May 2012, pp. 2330.
[225] D.-E. Spanos, P. Stavrou, N. Konstantinou, and N. Mitrou, Sensorstream: A semantic real-time stream management system, in International Journal of Ad Hoc and Ubiquitous Computing, 2011.
[226] T. Heath and C. Bizer, Linked Data (Synthesis Lectures on the
Semantic Web: Theory and Technology), 1st ed., J. Hendler, Ed.
454
[227]
[228]
[229]
[230]
[231]
Morgan & Claypool Publishers, February 2011. [Online]. Available:

http://linkeddatabook.com/editions/1.0/
D. L. Phuoc and M. Hauswirth, Linked open data in sensor data
mashups, in In Proc. 2nd International Workshop on Semantic
Sensor Networks (SSN09), vol. 522. CEUR Workshop at ISWC
2009, Washington DC, USA, 2009, pp. 116. [Online]. Available:
http://ceur-ws.org/Vol-522/p3.pdf
C. Perera, A. Zaslavsky, P. Christen, and D. Georgakopoulos,
Ca4iot: Context awareness for internet of things, in IEEE
International Conference on Conference on Internet of Things
(iThing), Besanon, France, November 2012, pp. 775782. [Online].
Available: http://dx.doi.org/10.1109/GreenCom.2012.128
C. Perera, A. Zaslavsky, P. Christen, M. Compton, and D. Georgakopoulos, Context-aware sensor search, selection and ranking
model for internet of things middleware, in IEEE 14th International
Conference on Mobile Data Management (MDM), Milan, Italy, June
2013.
E. Maximilien and M. Singh, A framework and ontology
for dynamic web services selection, Internet Computing, IEEE,
vol. 8, no. 5, pp. 84 93, sept.-oct. 2004. [Online]. Available:
http://dx.doi.org/10.1109/MIC.2004.27
S. Ran, A model for web services discovery with qos, SIGecom
Exch., vol. 4, no. 1, pp. 110, Mar. 2003. [Online]. Available:
http://doi.acm.org/10.1145/844357.844360
Charith Perera received his BSc (Hons) in Computer Science in 2009 from Staffordshire University, Stoke-on-Trent, United Kingdom and MBA in
Business Administration in 2012 from University
of Wales, Cardiff, United Kingdom. He is currently
pursing his PhD in Computer Science at The Australian National University, Canberra, Australia. He
is also working at Information Engineering Laboratory, ICT Centre, CSIRO and involved in OpenIoT
Project (Open source blueprint for large scale self
organizing cloud environments for IoT applications),
which is co-funded by the European Commission under seventh framework
program. His research interests include Internet of Things, pervasive and
ubiquitous computing with a focus on sensor networks, middleware, context
aware computing, mobile computing and semantic technologies. He is a
member of the Association for Computing Machinery (ACM) and the Institute
of Electrical and Electronics Engineers (IEEE).
Arkady Zaslavsky is the Science Leader of the
Semantic Data Management science area at Information Engineering Laboratory, ICT Centre, CSIRO.
He is also holding positions of Adjunct Professor
at ANU, Research Professor at LTU and Adjunct
Professor at UNSW. He is currently involved and
is leading a number of European and national research projects. Before coming to CSIRO in July
2011, he held a position of a Chaired Professor in
Pervasive and Mobile Computing at Lule University
of Technology, Sweden where he was involved in a
number of European research projects, collaborative projects with Ericsson
Research, PhD supervision and postgraduate education. Between 1992 and
2008 Arkady was a full-time academic staff member at Monash University,
Australia. Arkady made internationally recognised contribution in the area of
disconnected transaction management and replication in mobile computing
environments, context-awareness as well as in mobile agents. He made
significant internationally recognised contributions in the areas of data stream
mining on mobile devices, adaptive mobile computing systems, ad-hoc
mobile networks, efficiency and reliability of mobile computing systems,
mobile agents and mobile file systems. Arkady received MSc in Applied
Mathematics majoring in Computer Science from Tbilisi State University
(Georgia, USSR) in 1976 and PhD in Computer Science from the Moscow
Institute for Control Sciences (IPU-IAT), USSR Academy of Sciences in
1987. Before coming to Australia in 1991, Arkady worked in various research
positions at industrial R&D labs as well as at the Institute for Computational
Mathematics of Georgian Academy of Sciences where he lead a systems
software research laboratory. Arkady Zaslavsky has published more than 300
research publications throughout his professional career and supervised to
completion more than 30 PhD students. Arkady Zaslavsky is a Senior Member
of ACM, a member of IEEE Computer and Communication Societies.
Peter Christen is an Associate Professor in the

Research School of Computer Science at the Australian National University. He received his Diploma
in Computer Science Engineering from ETH Zurich
in 1995 and his PhD in Computer Science from
the University of Basel in 1999 (both in Switzerland). His research interests are in data mining and
data matching (entity resolution). He is especially
interested in the development of scalable and realtime algorithms for data matching, and privacy and
confidentiality aspects of data matching and data
mining. He has published over 80 papers in these areas, including in 2012
the book Data Matching (by Springer), and he is the principle developer of
the Febrl (Freely Extensible Biomedical Record Linkage) open source data
cleaning, deduplication and record linkage system.
Dimitrios Georgakopoulos is a Research Director

at the CSIRO ICT Centre where he heads the Information Engineering Laboratory that is based in Canberra and Sydney. The laboratory has 70 researchers
and more than 40 visiting scientists, students, and
interns specializing in the areas of Service/Cloud
Computing, Human Computer Interaction, Machine
Learning, and Semantic Data Management. Dimitrios is also an Adjunct Professor at the Australian
National University. Before coming to CSIRO in October 2008, Dimitrios held research and management
positions in several industrial laboratories in the US. From 2000 to 2008, he
was a Senior Scientist with Telcordia, where he helped found Telcordias
Research Centers in Austin, Texas, and Poznan, Poland. From 1997 to
2000, Dimitrios was a Technical Manager in the Information Technology
organization of Microelectronics and Computer Corporation (MCC), and the
Chief Architect of MCCs Collaboration Management Infrastructure (CMI)
consortial project. From 1990-1997, Dimitrios was a Principal Scientist at
GTE (currently Verizon) Laboratories Inc. Dimitrios has received a GTE
(Verizon) Excellence Award, two IEEE Computer Society Outstanding Paper
Awards, and was nominated for the Computerworld Smithsonian Award in
Science. He has published more than one hundred journal and conference
papers. Dimitrios is the Vice-Chair of the 12th International Semantic Web
Conference (ISWC 2013) in Sydney, Australia, 2013, and the General CoChair of the 9th IEEE International Conference on Collaborative Computing
(CollaborateCom 2013) in Austin, Texas, USA, 2013. In 2011, Dimitrios was
the General chair of the 12th International Conference on Web Information
System Engineering (WISE), Sydney, Australia, and the 7th CollaborateCom,
Orlando, Florida, October 2011. In 2007, he was the Program Chair of the
8th WISE in Nancy France, and the 3rd CollaborateCom in New York, USA.
In 2005, he was the General chair of the 6th WISE in New York. In 2002,
and he served as the General Chair of the 18th International Conference
on Data Engineering (ICDE) in San Jose, California. In 2001, he was the
Program Chair of the 17th ICDE in Heidelberg, Germany. Before that he
was the Program Chair of 1st International Conference on Work Activity
Coordination (WACC) in San Francisco, California, 1999, and has served as
Program Chair in a dozen smaller conferences and workshops.

Context Aware Computing For IoTs - Survey

Uploaded by

Copyright:

Available Formats

Context Aware Computing For IoTs - Survey

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Context Aware Computing For IoTs - Survey

Uploaded by

Copyright:

Available Formats

414

Context Aware Computing for

Manuscript received September 28, 2012; revised January 30, 2013.

of different aspects of context-aware computing, as we will

platform portability, context-awareness, security and privacy,

designed, developed and used in each project. This allows to

  

The IoT encompasses a significant amount of technologies

Fig. 2. Definition of the Internet of Things: The Internet of Things allows

and society. The magnitude of the applications can be seen in

communicate between themselves using wired and wireless

F. Layers in Sensor Networks

H. Characteristics of the IoT

Time considerations: The IoT could handle billions of

Fig. 4. Relationship between sensor networks and IoT.

acteristics of the IoT from a research perspective. Based on

I. Middleware Support for IoT

Synonyms [81]: Circumstance, situation, phase, position,

Legend: Device Management (DM), Interoperation (I), Platform

existing IoT focused solutions, which motivates us to survey

In addition, Sanchez et al. [85] explained the distinction

For example, the sensor readings produced by GPS sensors

2) Definition of Context-awareness: The term context

smart refrigerator [90] in the home to retrieve the shopping

      

  

Fig. 5. Context categorisation in two different perspectives: conceptual and

it could be identified as primary context. However, if we

consider more comprehensive categorisation schemes in a

detect a fire in a room in a smart home environment, the

Scalability and extensibility: The component should be

example, the justifications behind the results produced by the

Provide a broader guide that helps to identify the

Sensed, static, profiled,

Do not provide information about operational

Allow to model mental reasoning behind context

Do not provide information about core context,

Provide information about programming and coding

Weak in representing the relationship among

Provide information about context sources and the

More clear and structured method to organise context

Do not provide information about operational

Weak in representing the relationship among

event triggers an action autonomously in the IoT paradigm.

selected ten popular data life cycles to analyse in this survey.

1) Based on Responsibility: Context (e.g. sensor data)

Sensor hardware make the major decisions on sensing and

Decision on when to send data based on reasoning less

Software of the sensor data consumer makes the major

Can be used when sensors do not have knowledge on when to

Save energy due to no redundant network communications

More knowledge is required to identify the conditions and

Can be used to detect frost events or heat events in agricultural

without driver installations. In the IoT paradigm most objects

Either sensors can be configured to sense and communicate

usage, the term sensor is used to refer to tangible sensor

Direct Sensor Access

Efficient as it allows direct communication with the sensors