International Journal of Computer Applications (0975 – 8887)

National Conference on Advancements in Alternate Energy Resources for Rural Applications (AERA-2015)

Fog Computing
Pranati V. Patil
Computer science & Engg
College of Engg & Tech, Akola.

ABSTRACT
Fog Computing is a paradigm that extends Cloud computing
and services to the edge of the network. Similar to Cloud, Fog
provides data, compute, storage, and application services to
end users. In this article, we elaborate the motivation and
advantages of Fog computing, and analyses its applications in
a series of real scenarios, such as Smart Grid, smart traffic
lights in vehicular networks and software defined networks.
We discuss the state-of-the-art of Fog computing and similar
work under the same umbrella. Security and privacy issues are
further disclosed according to current Fog computing
paradigm. As an example, we study a typical attack, man-in-
the-middle attack, for the discussion of security in Fog
computing.

General Terms Fig.1. Fog between edge and cloud.

Cloud, Security, fog, privacy, Compute, Accountability.
2. NEED OF FOG
Keywords In the past few years, Cloud computing has provided many
Internet of things, Fog, cloud, smart grid. opportunities for enterprises by offering their customers a
range of computing services. Current “pay-as-you-go” Cloud
1. INTRODUCTION computing model becomes an efficient alternative to owning
CISCO recently delivered the vision of fog computing to and managing private data centers for customers facing Web
enable applications on billions of connected devices, already applications and batch processing .Cloud computing frees the
connected in the Internet of Things (IoT), to run directly at the enterprises and their end users from the specification of many
network edge. Customers can develop, manage and run details, such as storage resources, computation limitation and
software applications on Cisco IOx framework of networked network communication cost. However, this bliss becomes a
devices, including hardened routers, switches and IP video problem for latency-sensitive applications, which require
cameras. In Fog computing, services can be hosted at end nodes in the vicinity to meet their delay requirements. When
devices such as set-top-boxes or access points. The techniques and devices of IoT are getting more involved in
infrastructure of this new distributed computing allows people‟s life, current Cloud computing paradigm can hardly
applications to run as close as possible to sensed actionable satisfy their requirements of mobility support, location
and massive data, coming out of people, processes and thing. awareness and low latency. Fog computing is proposed to
Such Fog computing concept, actually a Cloud computing address the above problem. As Fog computing is implemented
close to the „ground‟, creates automated response that drives at the edge of the network, it provides low latency, location
the value. Cloud computing is the delivery of computing awareness, and improves quality-of-services (QoS) for
services over the Internet. Cloud services allow individuals streaming and real time applications[2]. Typical examples
and businesses to use software and hardware that are managed include industrial automation, transportation, and networks of
by third parties at remote locations. Examples of cloud sensors and actuators. Moreover, this new infrastructure
services include online file storage, social networking sites, supports heterogeneity as Fog devices include end-user
webmail, and online business applications. The cloud devices, access points, edge routers and switches. The Fog
computing model allows access to information and computer paradigm is well positioned for real time big data analytics,
resources from anywhere that a network connection is supports densely distributed data collection points, and
available. Cloud computing provides a shared pool of provides advantages in entertainment, advertising, personal
resources, including data storage space, networks, computer computing and other applications. Fog computing extends the
processing power, and specialized corporate and user paradigm to the edge of the network. while fog and cloud
applications. Both Cloud and Fog provide data, computation, using same resources(networking, computing, storage) and
storage and application services to end-users. However, Fog share many of the same mechanism and
can be distinguished from Cloud by its proximity to end-users, attribute(virtualization, multi-tenancy) the extension is a non
the dense geographical distribution and its support for trivial one in that there exist some fundamental differences
mobility. We adopt a simple three level hierarchy[2].Figure1. stemming from the reason fog computing developed: to
address and services that do not fit the paradigm of cloud[4].

1
 International Journal of Computer Applications (0975 – 8887)
National Conference on Advancements in Alternate Energy Resources for Rural Applications (AERA-2015)

3. FOG COMPUTING – A NATURAL 1. Low latency and location awareness;

BRIDGE BETWEEN CLOUD AND IoT 2. Wide-spread geographical distribution;
A new trend could be observed in joint relation with cloud – it
3. Wide area and real time Mobility access;
was the beginning of December 2013, discussing about a new
concept which behind exotic name founded the premises for 4. Increasing number and diversity of nodes;
distributed cloud models. In February 2015 we could describe
Fog Computing as essential bridge between the infinite power 5. Essential access to wireless;
of the Cloud and an almost infinite number of intelligent edge- 6. Strong presence of streaming and real time
points, conventionally conglomerated in IoT concept. applications,
7. Heterogeneity of devices and data sources.
The most illustrative example of Fog computing models in
real success stories for IoT are related to projects for
Connected Vehicles, Smart Grid, Smart Cities, Education,
Ecology, or Health Care. As described in the Cisco research
group article the Connected Vehicle deployment could
displays various connectivity scenarios: cars to cars, cars to
access points (Wi-Fi, 3G, LTE, roadside units [RSUs], smart
traffic lights), and access points to access points. The Fog has
a number of attributes that make it the ideal platform to
deliver a rich menu of SCV services in infotainment, safety,
traffic support, and analytics: geo-distribution (throughout
Fig.2. Bridge between fog and cloud.
cities and along roads), mobility and location awareness, low
What‟s happened with Fog Computing in this very short latency, heterogeneity, and support for real-time interactions.
period? Cloud computing confirmed most of tech and market
A smart traffic light system is based on smart traffic light
predictions, being involved in all major technology trends,
nodes, which interacts locally with a number of terrain sensors
from Big Data and Analytics, to Mobility, M2M, and Internet
which are detecting the presence of pedestrians and bikers,
of Everything… In parallel with this, IoT concept gains an
and measures the distance and speed of approaching vehicles.
explosive development, having as main engine exponential
It also interacts with neighboring lights to coordinate the green
increasing of data volumes provided by more and more
traffic wave. Based on this information the smart light sends
population and large variety of intelligent devices. In January
warning signals to approaching vehicles, and even modifies its
2014 Cisco reveled his own fog computing vision, designed
own cycle to prevent accidents.
on the idea of bringing cloud computing capabilities to the
edge of the network, much closer to the growing number of
user devices that are consuming cloud services and generating
3.1 A Internet of Things (IoT)
The Internet of Things (IoT) is a scenario in which objects,
the increasingly massive amount of data. Short time after,
animals or people are provided with unique identifiers and the
Cisco unveiled the company‟s IOx platform, designed to bring
ability to transfer data over a network without requiring
distributed computing to the network edge. According Cisco,
human-to-human or human-to-computer interaction. IoT has
Fog Computing extends the cloud computing paradigm to the
evolved from the convergence of wireless technologies,
edge of the network. While fog and cloud use the same
micro-electromechanical systems (MEMS) and the Internet.
computing, storage or network resources, and share many of
The Internet of Things (IoT, sometimes Internet of
the same mechanisms and attributes, like multi-tenancy and
Everything) is the network of physical objects or "things"
virtualization. Now we could speak about a Fog Computing
embedded with electronics, software, sensors, and
ecosystem based on Fog conceptually extension of Cloud
connectivity to enable objects to exchange data with the
computing – covering in a denser way wider geographic
manufacturer, operator and/or other connected devices based
locations, and concentrations of Fog devices – much more
on the infrastructure of International Telecommunication
heterogeneous in nature, ranging from end-user terminals,
Union's Global Standards Initiative. The Internet of Things
access points, to networks edge routers and switches. Provided
allows objects to be sensed and controlled remotely across
data should be processed locally in smart devices rather than
existing network infrastructure, creating opportunities for
being sent for processing in the cloud. Fog computing is one
more direct integration between the physical world and
approach to dealing with the demands of the ever-increasing
computer-based systems, and resulting in improved efficiency,
number small connected devices, sometimes referred to as the
accuracy and economic benefit. Each thing is uniquely
Internet of Things (IoT)[1].
identifiable through its embedded computing system but is
In the IoT scenario, a thing is any natural or man-made object able to interoperate within the existing Internet infrastructure.
that can be assigned an IP address and provided with the Experts estimate that the IoT will consist of almost 50 billion
ability to transfer data over a network. Some such things can objects by 2020.
create a lot of data. Mr. Michael Enescu offered as example a
jet engine, which can create 20 terabytes of engine
performance data in one hour. Sending all data to the cloud
and receiving the data back involve increasing demand for
bandwidth, considerable amount of time, and induced latency.
In a Fog Computing environment, a big part of local data
processing would take place in a router, rather than having to
be transmitted.
Resuming, the main characteristics of the Fog are:

2
 International Journal of Computer Applications (0975 – 8887)
National Conference on Advancements in Alternate Energy Resources for Rural Applications (AERA-2015)

Support for mobility Limited Supported

Real time Supported Supported
interaction
Types of last mile Leased line Wireless
connectivity

Table 2: Comparison of cloud and fog.

Cloud computing Fog computing
Data and applications are Rather than presenting and
processed in a cloud, working from a
which is time consuming centralized cloud, fog
task for large data. operates on a network
edge. So it consumes
less time.
Fig.3. Internet of things. Problems of bandwidth, as a Less demand of bandwidth,
result of sending every as every bit of data‟s
The term “Internet of Things” was coined by British bit of data over cloud were aggregated at
entrepreneur Kevin Ashton in 1999.Typically, IoT is expected channels. certain access point
to offer advanced connectivity of devices, systems, and instead of sending over
services that goes beyond machine-to-machine cloud channels.
communications (M2M) and covers a variety of protocols, Slow response time and By setting small servers
domains, and applications. The interconnection of these scalability problem as a called edge server in
embedded devices (including smart objects), is expected to result of depending visibility of the users, it
usher in automation in nearly all fields, while also enabling servers that are located is possible for fog
advanced applications like a Smart Grid and expanding to the at remote place. computing platform to
areas such as Smart city[1]. avoid response time and
scalability issues.
4. CLOUD VS FOG
The below tables show us the comparison between cloud and 5. ADVANTAGES AND
fog. As per the tables given below the latency and delay jitter
of the cloud computing is high but the fog computing is very DISADVANTAGES OF FOG
low. Location of services of cloud computing is within the COMPUTING
internet but of fog computing is at the edge of local network. 5.1 Advantages
Security in fog computing can be defined but in cloud 1. The detection of masquerade activity.
computing we can‟t defined it. Real time interactions are
supported by both cloud and fog computing. Data and 2. The confusion of the attacker and the additional
applications are processed in a cloud, which is time costs incurred to distinguish real from bogus
consuming task for large data. Rather than presenting and information.
working from a centralized cloud, fog operates on network
edge. So it consumes less time. There is a high probability of 3. The deterrence effect which although hard to
attack on data in cloud computing and fog computing having measure play a significant role in preventing
less probability of attack on data. In cloud computing slow masquerade activity by risk – averse attackers.
response time and scalability problems as a result of 4. Edge application services significantly decrease the
depending servers that are located at a remote place. By data volume that must be moved, the consequent
setting small servers called edge server in visibility of user, it traffic, and the distance the data must go, thereby
is possible for fog computing platform to avoid response time reducing transmission costs, shrinking latency, and
and scalability issues[3]. improving quality of service (QoS).
Table 1: Comparison of cloud and fog. 5. Edge computing eliminates, or at least de-
emphasizes, the core computing environment,
Requirements Cloud Fog computing
limiting or removing a major bottleneck and a
computing
potential point of failure.
Latency High Low
Delay jitter High Very low 6. Security is also improved as encrypted data moves
Location of service Within the At the edge of further in, toward the network core. As it
internet local approaches the enterprise, the data is checked as it
network passes through protected firewalls and other security
Distance between Multiple hopes One hope points, where viruses, compromised data, and active
server and hackers can be caught early on.
client
7. Finally, the ability to "virtualized" (i.e., logically
Security Undefined Can be defined
group CPU capabilities on an as-needed, real-time
Attack on data High probability Very low basis) extends scalability. The edge computing
probability market is generally based on a "charge for network
Location awareness No Yes services" model, and it could be argued that typical
Geo – Distribution Centralized Distributed customers for edge services are organizations
No. of server nodes Few Very large desiring linear scale of business application

3
 International Journal of Computer Applications (0975 – 8887)
National Conference on Advancements in Alternate Energy Resources for Rural Applications (AERA-2015)

performance to the growth of, e.g., a subscriber

base. 2. Smart Grids: Fog computing allow fast, machine to
machine (M2M) handshakes and human to machine
8. The significant reduction in data movement across interactions (HMI) which would work in cooperation
the network resulting in reduced congestion, cost with the cloud. Energy load balancing applications may
and latency, elimination of bottlenecks resulting run on network edge devices, such as smart meters and
from centralized computing systems, improved micro-grids Based on energy demand, availability and
security of encrypted data as it stays closer to the the lowest price, these devices automatically switch to
end user reducing exposure to hostile elements and alternative energies like solar and wind. As shown in
improved scalability arising from virtualized Figure, Fog collectors at the edge process the data
systems. generated by grid sensors and devices, and issue control
9. Edge computing, in addition to providing sub- commands to the actuators. They also filter the data to be
second response to end users, it also provides high consumed locally, and send the rest to the higher tiers for
levels of scalability, reliability and fault tolerance. visualization, real-time reports and transactional
analytics. Fog supports ephemeral storage at the lowest
10. Consumes less amount of band width[10][11]. tier to semi-permanent storage at the highest tier. Global
coverage is provided by the Cloud with business
a. Disadvantages intelligence analytics[1][7].
1. Nobody is identified when attack is happen.
2. It is complex to detect which user is attack.
3. We cannot detect which file was hacking[10].
Security is the biggest concern when it comes to fog
computing. By leveraging a remote cloud based infrastructure,
a company essentially gives away private data and
information, things that might be sensitive and confidential. It
is then up to the fog service provider to manage, protect and
retain them, thus the provider‟s reliability is very critical. A
company‟s existence might be put in jeopardy, so all possible
alternatives should be explored before a decision. On the same
note, even end users might feel uncomfortable surrendering
their data to a third party.

11. APPLICATION OF FOG Fig.5. Smart Grid.

COMPUTING
Following are the some application of fog computing. 3. Health care: The cloud computing market for healthcare is
expected to reach $5.4 billion by 2017, and fog computing
1. Connected Cars: Fog computing is the ideal for the would allow this on a more localized level.
connected vehicles(CV) because real time interactions will
make communication between cars, access points and traffic 4. Wireless Sensor and Actuator Networks: Traditional
lights as safe and efficient as possible. Video camera that wireless sensor networks fall short in applications that go
senses an ambulance flashing lights can automatically change beyond sensing and tracking, but require actuators to exert
street lights to open lanes for the vehicle to pass through physical actions like opening, closing or even carrying
traffic. Smart street lights interact locally with sensors and sensors. In this scenario, actuators serving as Fog devices can
detect presence of pedestrian and bikers, and measure the control the measurement process itself, the stability and the
distance and speed of approaching vehicles. As shown in oscillatory behaviors by creating a closed-loop system. For
Figure, intelligent lighting turns on once a sensor identifies example, in the scenario of self-maintaining trains, sensor
movement and switches off as traffic passes. Neighboring monitoring on a train‟s ball-bearing can detect heat levels,
smart lights serving as Fog devices coordinate to create green allowing applications to send an automatic alert to the train
traffic wave and send warning signals to approaching vehicles. operator to stop the train at next station for emergency
Wireless access points like Wi-Fi, 3G, road-side units and maintenance and avoid potential derailment. In lifesaving air
smart traffic lights are deployed along the roads. Vehicles-to- vents scenario, sensors on vents monitor air conditions
Vehicle, vehicle to access points, and access points to access flowing in and out of mines and automatically change air-flow
points interactions enrich the application of this if conditions become dangerous to miners[1][7].
scenario[5][4]. 5. Smart Cities: Fog computing would be able to obtain sensor
data on all levels of activities of cities and integrate all the
mutually independent network entities within. The
applications of this scenario are facilitated by wireless sensors
deployed to measure temperature, humidity, or levels of
various gases in the building atmosphere. In this case,
information can be exchanged among all sensors in a floor,
and their readings can be combined to form reliable
measurements. Sensors will use distributed decision making
and activation at Fog devices to react to data. The system
components may then work together to lower the temperature,
inject fresh air or open windows. Air conditioners can remove
Fig.4. Connected Cars. moisture from the air or increase the humidity. Sensors can

4
 International Journal of Computer Applications (0975 – 8887)
National Conference on Advancements in Alternate Energy Resources for Rural Applications (AERA-2015)

also trace and react to movements (e.g, by turning light on or 2) Compute/Storage limitation Current trends are improving
off). Fog devices could be assigned at each floor and could this fact with smaller, more energy-efficient and more
collaborate on higher level of actuation. With Fog computing powerful devices (e.g. one of today's phones is more powerful
applied in this scenario, smart buildings can maintain their than many high end desktops from 15 years ago). Still new
fabric, external and internal environments to conserve energy, improvements are granted for non consumer devices.
water and other resources[2].
3) Management In addition to setting up the communication
routes across end nodes, IoT/ubiquitous computing nodes and
applications running on top need to be properly setup and
configured to operate as desired. Having potentially billions of
small devices to be configured, the fog will heavily rely on
decentralized (scalable) management mechanisms that are yet
to be tested at this unprecedented scale. One thing that can be
predicted with certain degree of confidence is that there will
be no full control of the whole fog and asymptotic declarative
configuration techniques will become more common.
4) Security The same security concerns that apply to current
virtualized environments can be foreseen to affect fog devices
hosting applications. The presence of secure sandboxes for the
execution of droplet applications poses new interesting
challenges: Trust and Privacy. Before using other devices or
mini-clouds in the network to run some software, isolation and
sandboxing mechanisms must be in place to ensure
bidirectional trust among cooperating parties. The fog will
Fig.6. Smart Cities.
allow applications to process users data in third party's
12. PRIVACY hardware/software. This of course introduces strong concerns
Today, we constantly leak personal information by using about data privacy and its visibility for those third parties[6].
different products, platforms and services. We may think were 5) Standardization Today no standardized mechanisms are
in charge of our shopper cards and our mobile apps and our available so each member of the network (terminal, edge
smart fridges, but lets not fool ourselves. The information is point...) can announce its availability to host others software
not ours. It belongs to Google, and IBM, and Cisco Systems components, and for others to sent it their software to be run.
and the global Mega Corp that owns your local supermarket.
If you don‟t believe us, just try removing your data from their 6) Accountability/Monetization Having users able to share
databases. Users are becoming increasingly concerned about they spare resources to host applications is crucial to enable
the risk of having their private data exposed. As a result, new business models around the concept of the fog. A proper
besides the technical challenges introduced by the ubiquity of system of incentives needs to be created. The incentives can
devices, there is another trend that will push for a fog scenario financial or otherwise (e.g. unlimited free data rates). On the
where data is not sent to a few centralized services, but it is other hand the lack of central controlling entity in the fog
instead kept 'in the network' for better privacy. Data makes it difficult to assert if a given device is indeed hosting a
ownership will be a very important cornerstone of the fog, component (droplet) or not.
where some applications will be able to use the network to run 7) Programmability Controlling application lifecycle is
applications and manage data without relying on centralized already a challenge in cloud environments. The presence of
services.. Storing encrypted sensitive data in traditional clouds small functional units (droplets) in more locations (devices)
is an alternative to keep privacy. However, this makes it really calls for the right abstractions to be in place, so that
hard to perform any processing over such data. There is programmers do not need to deal with these difficult issues.
important research work on this topic, for example using Easy to use APIs for programmers will heavily rely on simple
crypto-processors or applying special encryption functions Management mechanisms that provide them with the right
that cipher while keeping some of its original properties, thus abstractions to hide the massive complexity of the fog. Some
allowing to perform certain limited tasks on it. Still, such vendors like Microsoft have already taken some steps in
options have limited applicability. As a result, users will positioning themselves in this space9[8].
demand innovative ways to preserve their privacy from any
potential big-brother-like entity. This will be a great incentive 14. CONCLUSION
to adopt fog technologies, as they will enable the network to The fog is nothing but the convergence of a set of
replace centralized services[3][9]. technologies that have been developing and maturing in an
independent manner for quite some time. The integration of
13. CHALLENGES AHEAD these into a single IT scenario is an answer to the new
There are many open problems that will have to be addressed requirements introduced by device ubiquity and demands for
to make the fog a reality. It is necessary to clearly identify agile network and service management and data privacy. As a
them so future research works have these problems into result the fog will dramatically shift many of our current
account. The set of open challenges for the fog to become a practices at almost every layer of the IT stack, like apps
reality is: development, network traffic management, network/service
1) Discovery/Sync Applications running on devices may need provision, accounting, apps collaboration mechanisms, etc.
either some agreed centralized point (e.g. establish an This article has provided a broad overview of this
upstream backup if there are too few peers in our storage convergence and what are the common points that link all
application). these technologies together, creating a new paradigm that
some have already named as fog computing.

5
 International Journal of Computer Applications (0975 – 8887)
National Conference on Advancements in Alternate Energy Resources for Rural Applications (AERA-2015)

15. ACKNOWLEDGMENTS [5] F. Bonomi, “Connected vehicles, the internet of things,

It is a matter of great pleasure to highlight a fraction of and fog computing,” in The Eighth ACM International
knowledge, I acquired during my technical education through Workshop on Vehicular Inter- Networking (VANET),
this seminar. This would not have been possible without the Las Vegas, USA, 2011.
guidance and help of many people. This is where I have the [6] Salvatore J. Stolfo,” Fog Computing: Mitigating Insider
opportunity of expressing gratitude from the core of my heart. Data Theft Attacks in the Cloud”, Computer Science
Thanks are in order to all the colleagues and friends who Department, Columbia University New York , NY, USA.
knowingly or unknowingly helped me during this work.
[7] Tom H. Luan, Longxiang Gao, Yang Xiang,” Fog
16. REFERENCES Computing: Focusing on Mobile Users at the Edge”,
[1] F. Bonomi, R. Milito, P. Natarajan. “Fog computing School of Information Technology Deakin University,
– A platform For Internet Of Things And Analytics” Burwood Melbourne, VIC 3125, Australia
Enterprise Networking Labs, Cisco Systems Inc., San
[8] Luis M. Vaquero Hewlett-Packard Labs Bristol, United
Jose, USA
Kingdom,luis.vaquero@hp.com
[2] Ivan Stojmenovic, “The fog computing Paradigm :
[9] W. Wang and Z. Lu, “Survey cyber security in the smart
Scanerios And Security Issues”, SIT, Deakin University,
grid: Survey and challenges,” Comput. Netw., vol. 57, no.
Burwood, Australia and SEECS, University of Ottawa,
5, pp. 1344–1371, Apr. 2013.
Canada.
[10] http://www.slideshare.net/saisharansai/fog-computing-
[3] Luis. Rodero-Merino,” Finding your Way in the Fog:
46604121
Towards a Comprehensive Definition of Fog
Computing”, HP LaboratoriesHPL-2014-60 [11] http://www.slideshare.net/professorbanafa/what-is-fog-
[4] Sheng Wen,” The Fog Computing Paradigm: Scenarios computing
and Security Issues”, School of Information Technology,
Deakin University, 220 Burwood Highway, Burwood,
VIC, 3125, Australia.

IJCATM : www.ijcaonline.org 6