This article has been accepted for publication in a future issue of this journal, but has not been

fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2017.2735988, IEEE Access

LEO Satellite Constellation for Internet of

Things
Zhicheng Qu, Gengxin Zhang, and Jidong Xie, Member, IEEE


TABLE I
Abstract—Internet of Things (IoT) is one of the evolutionary
directions of the Internet. This paper focuses on the low earth CONNECTION TYPES UNDER DIFFERENT SCENARIOS
orbit (LEO) satellite constellation-based IoT services for their
irreplaceable functions. In many cases, IoT devices are Connection
Scenario Devices
distributed in remote areas (e.g. desert, ocean, forest, etc.); type
especially in some extreme topography where direct access to Health monitoring Short-range,
terrestrial network is impractical, they can only be covered by E-health
devices cellular
satellite. Comparing to traditional geostationary earth orbit
(GEO) systems, LEO satellite constellation takes advantages of Home Smart household
Short-range
low propagation delay, small propagation loss and global automation electrical appliances
coverage. Furthermore, revision of existing IoT protocol are Vehicles, lighting,
necessary to enhance the compatibility of the LEO satellite monitoring cameras,
constellation based IoT with terrestrial IoT systems. In this work, Smart city Cellular
smart parking devices,
we provide an overview of the architecture of the LEO satellite
air quality monitor
constellation-based IoT including the following topics: LEO
satellite constellation structure, efficient spectrum allocation, Smart grid Power meters LPWAN
heterogeneous networks compatibility, and access and routing Environmental Wild environmental
protocols. LPWAN
monitoring sensors
Geologic
Index Terms—Internet of Things (IoT), LEO satellite Crustal movement
constellation, Low-Power Wide-Area Network (LPWAN), long
disaster LPWAN
monitor
range (LoRa), machine-to-machine (M2M) communications, forecasting
Narrow Band Internet of Things (NB-IoT).
business management, environmental monitoring and smart
city [1-3]. Connection types under different IoT scenarios are
I. INTRODUCTION listed in Table I.
Machina Research, a global leading provider of market
T HE Internet of Things (IoT) is a burgeoning paradigm that
points out a novel direction of future internet, in which
numerous heterogeneous networks containing different user
intelligence and strategic, predicts 27 billion connected devices
and USD 3 trillion in revenue in 2025. It highlights the
data will be integrated transparently and seamlessly through direction of connections using Low Power Wide-Area Network
appropriate protocol stacks [1], [2]. This integration aims to (LPWAN) [4]. Comparing to short-range connections based on
enable anything with a transceiver to access Internet at all times WiFi, Zigbee, Near Field Communication (NFC), Bluetooth or
and places. Moreover, through those easy accesses, various in-building power line communication (PLC) and cellular
kinds of IoT devices such as, environmental monitoring sensors, connections, the wide range ones are more capable for remote
smart household electrical appliances, actuators, vehicles, etc., industrial scenario like smart grid and environmental
can exchange data with the IoT networks and provide monitoring. However, terrestrial-based LPWAN still cannot
unprecedented services for private users, business users, cover remote areas like desert, coastal waters and forests due to
government users, army and anyone who utilize IoT. The the commercial and engineering difficulties of constructing
obvious applications of the IoT will be visible for individual LPWAN in those areas.
needs such as e-health, home automation, and elderly To solve the aforementioned covering problem, satellite
assistance as well as for industrial needs like smart grid, communication for IoT comes into view. The potential
necessities of constructing satellite IoT system are listed as
follows:
Z. C. Qu is with Department of Communication Engineering, Nanjing
1) Firstly, the extreme topographies, such as cliff, valley, and
University of Posts and Telecommunication, Nanjing, Jiangsu Province 210003, steep slope, are places where geologic disasters are more
China. (e-mail: 1015010132@njupt.edu.cn) easily to happen while terrestrial networks cannot access
G. X. Zhang and J. D. Xie are with Department of Communication due to engineering difficulties. Satellite IoT system can
Engineering, Nanjing University of Posts and Telecommunication, Nanjing,
Jiangsu Province 210003, China. make a breakthrough in the limits of topography with its
Corresponding author: Gengxin Zhang (e-mail: satlab@126.com)

2169-3536 (c) 2017 IEEE. Translations and content mining are permitted for academic research only. Personal use is also permitted, but republication/redistribution requires IEEE permission. See
http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2017.2735988, IEEE Access

covering advantages. of long term evolution (LTE) by deleting physical uplink

2) For IoT application in remote areas, satellite IoT system control channel (PUCCH) [9], [10], narrowing the channel
provides a cost-efficient solution with respect to other bandwidth [11], and predigesting the process. The core of
terrestrial technologies to their interconnection and NB-IoT, however, persists in the form of LTE, which requires
communication with “the rest of the world” [5], [6]. uninterrupted connection between user equipment (UE) and the
3) For terrestrial IoT network, which mostly depends on Evolved Node B (eNodeB) during the whole communication
wireless access, a communication network consisting of process. If such protocol are deployed in satellite IoT system
enough base stations is indispensable. However, with its intact type, considering the limited communication
constructing terrestrial base stations and connecting resources over satellite and required low power consumption
network is constrained by several limitations. For instance, for terminals, the interactive overhead is a phenomenal burden
terrestrial communication infrastructures are fragile that that SBD services and satellite IoT terminals cannot afford.
may be easily damaged by natural disasters like In this work, we particularly focused on the following issues:
earthquake and flood. Meanwhile, terrestrial IoT can 1) design of LEO satellite constellation for IoT services; 2)
provide effective coverage only in a relatively small range refinement of terrestrial IoT protocols for suiting satellite
(currently, terrestrial wireless network can only cover communication; and 3) anti-interference measures for satellite
around 20% of territory in China and the U.S [7]). As a IoT system.
supplement and extension to the terrestrial IoT network,
satellite IoT system is the only approach to achieve global II. RELATED WORKS
IoT service covering. Satellite-based systems, especially LEO satellite systems, have
Furthermore, aiming at the necessities of satellite IoT system, been adopted in machine-to-machine (M2M) communications
low earth orbit (LEO) satellite constellation technology has (e.g., [12-14]) as well as data collecting systems [15] for years.
unique advantages over geostationary earth orbit (GEO) In [12], an architecture for a M2M satellite system relies on a
systems: small LEO constellation has been proposed. In [13], the authors
1) Due to the lower orbit altitude of LEO satellite have systematically evaluated the proposed architecture for
constellation (normally lower than 2000km), it is more satellite-based M2M networks in terms of clustering,
time efficient than GEO systems. In terms of propagation connectivity and availability. [14] is the homepage of
delay, quantified by a round trip time (RTT), LEO satellite “ORBCOMM”, which is a leading global provider of M2M
constellation has a RTT less than 100ms while GEO utilizing LEO satellite constellation (orbit altitude: 815km). In
systems’ RTT is over 600ms [6]. [15], the authors discussed services providing by
2) Most satellite IoT terminals are designed to be small-sized, satellite-routed sensor system (SRSS) and focused on a novel
long-life, and low power consumption. Benefiting from bandwidth allocation method. In the aforementioned proposals,
the relatively shorter propagation distance of LEO satellite the structure of LEO constellation system providing M2M
constellation, the signal loss due to propagation shall be services has been clarified. Furthermore, in [6] and [16], the
authors proposed the conception of satellite-based IoT.
smaller, which helps the terminal design to reach the ideal
Moreover, the authentication and security issues in IoT field
pattern.
have been studied among numerous application scenarios,
3) Communication via GEO satellite is constrained by
including E-health and smart home (e.g., [17-20]). In [18], the
extreme topographies because of the relatively static authors proposed an IoT based cloud centric structure
position between terminals and GEO satellite. If there is providing user activities analysis, and evaluating its efficiency
an obstacle (tree, cliff, etc.) in the line of sight (LOS) from and security. In [19] and [20], device fingerprinting and logical
terminal to satellite, this terminal is unable to sensing are integrated in security design of smart home.
communicate with the satellite unless the obstacle is Although LEO constellation networks have been studied
removed. Comparing to GEO satellite, LEO satellite is from many aspects, including resource allocation (e.g., [15]),
connectable even if an obstacle locates near the terminal IPv6 over satellite (e.g., [21]), and QoS mapping in
due to the satellite movement. heterogeneous satellite-based networks (e.g., [22]), few studies
In fact, LEO satellite constellation-based IoT system is a have addressed on the compatibility of higher layer protocols
powerful supplement to the terrestrial IoT networks. The global between satellite-based and terrestrial IoT systems. As
market for satellite IoT services is going to reach 1.7 billion mentioned before, satellite-based IoT systems are concerned as
dollars in 2017, and will rapidly dilate in coming years [8]. the supplement and extension of terrestrial ones, which requires
However, the IoT services requirements cannot be fully good compatibility in the whole protocol stacks. Moreover,
satisfied by simply combining the current satellite proprietary studies focusing on the feasibility of satellite-based IoT
standard and terrestrial IoT protocol stacks. For example, systems are insufficient. The application of LEO satellites for
plenty of IoT services tend to be short-burst data (SBD) dense networks of M2M sensor devices deployed in several
transmission. In terrestrial IoT systems, based on complete geographical locations is proposed in [13]. However, the
authors considered the employment of Ka band, which is
access networks and high base stations density, applications
difficult to use in IoT scenario since that operating at Ka band
can inherit the wireless communication protocol with a little
requires larger antenna scale and power consumption, which is
refinement to suit their requirements. Narrow Band Internet of
contrary to the demands of IoT UE (small-sized and low power
Things (NB-IoT) protocol, for instance, simplifies the structure

2169-3536 (c) 2017 IEEE. Translations and content mining are permitted for academic research only. Personal use is also permitted, but republication/redistribution requires IEEE permission. See
http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2017.2735988, IEEE Access

characteristics of frequent and prolonged temporary

disconnections, and long propagation delay [24]. One of the
typical DTAs is water monitoring.
Water is the source of life, and over 70 percent of earth
surface is water covered. Water monitoring aims to secure the
water quality from human activities and human safety from
natural disasters. The main category of water monitoring
includes temperature monitoring, tide monitoring, pollution
monitoring, etc. The use of satellite could be irreplaceable for
remote waters monitoring (for example, wetland and ocean),
which terrestrial system cannot make implementation.
Currently, satellite-based monitoring remains in the way of
remote sensing including satellite imagery [25], on-board
satellite sensors[26], and on-board satellite Synthetic aperture
radar (SAR) systems [26], [27]. Those traditional techniques
Fig. 1. Satellite-based DTN system. have limits listed as follows:
1) Weather influence: Aforementioned sensing techniques,
consumption). Besides, in [12], the authors did the calculation especially the optical sensing by satellite imagery, are
of link budget without considering interference from terrestrial impacted by weather conditions. Fog, cloud or brume will
systems operating on the same frequency band. It should be lead to inaccuracy of satellite imagery. Meanwhile, when
highlighted that the co-frequency interference suffered by the comparing selected data on the timeline, different weather
LEO satellite is the critical factor to determine the system’s conditions will eliminate the comparability.
feasibility, and anti-interference measures are needed in the 2) Indirect results: The results from sensing need to be
system design. In this paper, we mainly focus on the IoT analyzed by specialists to get underlying information. This
features combining with the LEO satellite characteristics, and indirect approach will increase the difficulty of using
the main contributions of this paper are summarized as follows: analyzing algorithms as well as reduce analyzing
1) A united higher layer design for integrated space-terrestrial efficiency.
(IST) IoT network is presented. 3) System cost: Usually, remote sensing satellites are
2) A co-frequency interference analysis under different designed for specific purpose. Therefore, to gather
covering circumstances is offered, and potential different kinds of water information, the whole system
anti-interference measures are proposed. needs to launch corresponding satellites, which certainly
The reminder of this paper is organized as follows. Section will increase the constructing and operating costs.
III describes several typical application scenarios for LEO Aiming at the above disadvantages of remote sensing, LEO
satellite constellation-based IoT system. Section IV presents constellation-based IoT system may offer a replaceable
the design consideration on LEO satellite constellation. Section solution for water monitoring, as it would obtain direct
V demonstrates the compatibility between satellite-based and monitoring information through different kinds of sensors. In
terrestrial systems. Section VI illustrates existing interference this solution, LEO satellites only play the role of
environment between terrestrial and satellite-based network, communication platform, which can sharply reduce the cost of
and demonstrates potential anti-interference measures for a single satellite. Furthermore, LEO constellation can ensure
satellite IoT system. Finally, Section VII concludes this work. more frequent data-collection than a single remote sensing
satellite to improve the accuracy of prediction and forecasting.
III. TYPICAL APPLICATION SCENARIOS FOR LEO SATELLITE Though LEO constellation seems an ideal plan for water
CONSTELLATION-BASED IOT SYSTEM monitoring, several challenges arise. For example, to meet the
In this section, we divide typical LEO satellite constellation- requirement of energy efficient, it requires specific concern on
based IoT application scenarios into two groups: 1) protocols and media access control (MAC) mechanism.
delay-tolerant applications (DTAs) (for instance, monitoring Moreover, interference from terrestrial IoT systems will affect
and forecasting applications); 2) delay-sensitive applications the monitoring performance, especially in wetlands or coastal
(DSAs) (for instance, enhanced supervisory control and data waters. Those issues are discussed in Section V and Section VI.
acquisition (SCADA) and military applications). For each B. Delay-Sensitive Applications
group, we present one or two specific applications to illustrate
the current capabilities and potential possibilities which LEO DSAs are quite different scenarios that have stringent
constellation-based IoT system owns. requirements (i.e., lower latency and higher reliability) from the
DTAs. Smart grid and Internet of Battle Things (IoBT) are
A. Delay-Tolerant Applications representative application scenarios for civil and for military,
The concept of DTA is a part of delay-tolerant networking respectively.
(DTN), which is a novel communication structure to provide Current grid, for instance, is under the SCADA scheme, in
automated store-and-forward data communication services in which remote monitoring and automated control of substations
networks [23]. Fig. 1 shows an overview of a satellite-based are implemented through a slow central network [28]. However,
DTN system. In general, those applications have common the novel concept of smart grid [29] requiring the power grid to

2169-3536 (c) 2017 IEEE. Translations and content mining are permitted for academic research only. Personal use is also permitted, but republication/redistribution requires IEEE permission. See
http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2017.2735988, IEEE Access

TABLE II
ORBIT PLANE SATELLITE PARAMETERS FOR LEO CONSTELLATION WITH ISLS

Perigee Apogee Right Ascension of Argument of

Orbit Plane Inclination True Anomaly
Altitude/km Altitude/km Ascending Node Perigee
A 887 887 86º 0º 0º 0º
B 887 887 86º 38º 0º 333º
C 887 887 86º 76º 0º 353º
D 887 887 86º 114º 0º 328º
E 887 887 94º 151º 0º 351º

be able to react and adapt to the grid dynamics can be defined as

a DSA. Currently, parts of smart grid elements are already
available, and existed wired/wireless communication networks
can support smart grid in urban/suburban areas. Apart from the
densely populated regions, for implementing smart grid in
remote locations including offshore wind farms and solar
energy systems in desert, LEO constellation-based IoT system
could present a viable and cost-effective solution. In [30], a
LEO satellite constellation-based power manage solution is
illustrated in detail including LEO satellite network, delay
analysis and simulation experiments for typical traffic
scenarios, which cover the main concerning points of smart Fig. 2. 2-D LEO constellation without ISLs coverage diagram.
grid.
Modern war is mainly in the form of information-based war,
and United States Forces proposed the concept of network
centric warfare (NCW) [31]. The key of NCW is that things can
communicate with each other and can better serve humans
involved in warfare. Similarly, the intelligent devices that
populate in warfare is referred to as the IoBT [32]. During
warfare, ground access system would be vulnerable as a result
of enemy actions so that satellite-based access system is of
great significance. However, GEO satellite cannot meet the
safety requirements of IoBT for different reasons:
1) GEO satellite is relatively static to the ground, which makes
it easy to be located and blanket jammed by enemies.
2) GEO satellite is easy for signal tracking due to its wide beam
coverage. Once characteristic parameters have been seized,
using deception jamming or coherent jamming will
significantly raise the interference level.
Due to the aforementioned drawbacks of GEO satellite in
IoBT, LEO constellation based system becomes a better choice.
Meanwhile, in IoBT region, DSAs such as unmanned combat
robot and unmanned aerial vehicle can be only satisfied by Fig. 3. 3-D LEO constellation without ISLs coverage diagram.
LEO constellation based network due to the strict requirements 2) Target application scenarios
on latency. 3) Cost of single satellite and constructing the whole
Comparing to DTAs, DSAs require different design in LEO constellation
satellite constellation, e.g. implementing inter-satellite links Aiming at above factors, LEO constellation can be
(ISLs). At the same time, satellites are required to have the categorized into two classes, i.e. constellation with/without
ability of on-board processing and on-board routing. The ISLs. Meanwhile, there are differences in network architecture
detailed demonstrations are illustrated in Section IV and characteristics between two constellation designs.
Section V.
A. LEO Constellation without ISLs
IV. DESIGN OF LEO SATELLITE CONSTELLATION LEO constellation without ISLs is more suitable for DTAs due
Generally, a LEO satellite constellation consists numbers of to its low cost and complexity. At the beginning of designing a
satellites in orbits of 500-2000 km. In constellation designing, constellation, the orbit eccentricity, altitude and inclination
following main factors should be taken into consideration: should be considered.
1) Global coverage

2169-3536 (c) 2017 IEEE. Translations and content mining are permitted for academic research only. Personal use is also permitted, but republication/redistribution requires IEEE permission. See
http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2017.2735988, IEEE Access

, e.g. orbit period is one day and recursive cycles

are 14, to calculate out and .
To realize global seamless covering, Rosette Constellations
are recommended due to their coverage properties [33].
Commonly, a shorthand notation named Walker
code is used to designate a rosette constellation having total
satellites, orbit planes, and a harmonic factor . In general,
Fig. 4. 2-D LEO constellation with ISLs coverage diagram.
reasonable inclination of orbit is supposed to range from
to for the covering goal. In this work, a Rosette
Constellation which inclination of orbit is and walker code
(35, 7, 1) is proposed. Meanwhile, in order to optimize polar
area coverage, the whole constellation design adds two polar
satellites to cover polar areas periodically. The 2-D and 3-D
coverage diagrams are shown in Fig.2 and Fig. 3 by simulating
in System Tool Kit (STK).
B. LEO Constellation with ISLs
To implement ISLs and realize global seamless coverage, LEO
constellation tends to accept polar orbit planes instead of
inclining orbit planes. However, to avoid satellites crashing at
poles when there are two or more polar orbit planes, near polar
orbits with inclination ranging from to (except )
are used to replace the standard polar orbit. Meanwhile, to
reduce constructing cost, the constellation needs minimum
satellites to realize global coverage. ISLs between orbit planes,
Fig. 5. 3-D LEO constellation with ISLs coverage diagram. which dramatically increase the system complexity (for
instance, on-board processing, satellite antenna pointing, and
As an important parameter in orbit designing, orbit system routing), are ignored by common LEO constellation
eccentricity will have influence on satellite’s covering area and based IoT system for same reason. To be notified, in IoBT,
time. When the satellite is near the apogee, covering area and ISLs between different planes and on-board routing are
time becomes relatively wider and higher, respectively. On the necessary for strictly real-time application. Therefore, method
contrary, when the satellite is near perigee, the two quotas will of combining coverage zones of different orbit planes is used to
decrease at same time. To get constant satellite overhead pass design LEO constellation. The orbit planes consisted in the
times and power levels needed for communication, the satellite constellation have the same orbit altitude, and the satellites in
orbits in the constellation are all round orbit. Meanwhile, in each plane have the same inclination and angular spacing.
order to be convenient for controlling satellites during Here, a constellation design with 40 satellites in five orbit
operation, the orbits are designed as quasi-recursive orbit, e.g. planes is proposed. In this plan, each satellite has two
satellites will pass the same point after a certain time interval in bidirectional ISLs with its two vertically adjacent satellites in
days. Therefore, the satellite orbit period can be calculated the same plane. The initial orbit parameters, 2-D, and 3-D
by following equations: coverage diagrams are shown in Table II, Fig.4, and Fig.5,
respectively.
(1)
V. Compatibility between LEO Constellation-Based and
(2) Terrestrial IoT Systems
As mentioned at the beginning, LEO constellation-based IoT
Equation (1) ensures that the orbit is quasi-recursive, where system is a powerful supplementary to terrestrial system, which
is equinoctial day with length of 86164 s, and , are aims to cover remote or extreme areas where terrestrial system
integers, which represent orbit period in days and recursive cannot reach. Therefore, the compatibility between two
cycles, respectively. Equation (2) is the Kepler's third law, in systems should be considered in order to generate an integrated
IoT network. Unfortunately, though the third generation
which is the Kepler constant defined by partnership project (3GPP) LTE air interface in satellite
, is orbit altitude, and is the communication was evaluated in [34], up to now, the terrestrial
radius of earth with 6371 km. In this work, we pick IoT standards have not taken any satellite components into
consideration. Hence, the compatibility remains an open area
for researching.

2169-3536 (c) 2017 IEEE. Translations and content mining are permitted for academic research only. Personal use is also permitted, but republication/redistribution requires IEEE permission. See
http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2017.2735988, IEEE Access

and IoT terminals in other satellites’ footprints without

terrestrial infrastructures’ support.
Corresponding to the two satellite design approaches, the
LEO constellation-based IoT network can be depicted by
EGS-centralized network and dynamic satellite topology
network shown in Fig. 6 and Fig. 7, respectively. To be noticed
EGS
footprint j
that EGS-centralized network mainly cater to DTA and utilize
footprint i
constellation without ISLs. On the contrary, dynamic satellite
Terrestrial
topology mostly aims on DSA and adopts ISLs in constellation
network
design.
IoT mostly depends on TCP/IP protocol suite where IP
Fig. 6. EGS-centralized network architecture.
ISL
provides enough routing information and TCP guarantees
ISL
reliable data transmission on top of IP. To ensure the
ISL
LEO compatibility between LEO constellation-based IoT network
constellation
and terrestrial ones, and distinguish from M2M
communications, TCP/IP should be implemented in the
network design. Meanwhile, since the network architecture or
topology is decided by satellite design, the routing strategies
vary from one to another.
footprint i
EGS
footprint j For EGS-centralized network, satellites are not always
visible for all IoT terminals, which makes it difficult to design
Terrestrial routing strategy. For instance, in Fig. 6, and are
network
corresponding to footprint and , respectively. Since and
Fig. 7. Dynamic satellite topology network architecture. may not be visible for EGS, routing to footprint or need
to wait for the satellite recurrence. To deal with this problem, a
The aspect of compatibility includes the following topics proxy cache scheme can be implemented in the routing strategy,
such as MAC protocols, network architecture and united which EGSs store all routing information (e.g. terminals
service patterns. In literature, MAC protocols for location and ephemeris) updated within each transmission
satellite-based IoT system have been studied mainly focus on process. When EGS receive a routing request, the route via an
random access (RA) in the case where numerous terminals are optimum satellite to the destination will be calculated basing on
deployed and the transmitting requests are generated arbitrarily. ephemeris. The recursive orbit also designed for this routing
In [6], [16, 35, 36], different RA protocols are analyzed for their strategy to get relatively stable routing table. Moreover, proxy
feasibilities in satellite-based IoT system. For instance, in [16], cache scheme is also considerable in united higher layer
satellite uses a divide-and-conquer scheme to allocate time designing, which will be detailed in Section V-B.
slots on demand to terminals under the assumption of using Routing situation becomes more complex in dynamic
GEO satellite which means the whole system is time satellite topology network (DSTN) due to the relative motion
synchronous. However, in LEO constellation-based IoT system, between each satellite. The topology is a variation of the
high satellite dynamic makes system time synchronous difficult Manhattan street network (MSN) [38], which is a regular,
to realize due to the dramatic relative motion between satellite two-connected network, designed for packet communications.
and UEs. Therefore, to adopt those RA schemes in LEO The fundamental difference in routing strategy of MSN from
constellation-based IoT system potentially, satellite dynamic conventional loop network is that routing decisions must be
problem needs to be mitigated. made at every node (e.g., each satellite in LEO constellation
The following of this section focuses on the other two topics: network) in this network [39]. However, unlike terrestrial
1) network architecture of LEO constellation-based IoT and 2) networks, where a fixed route between source and destination is
the united higher layer design for integrated space-terrestrial available at any point in time, LEO ISL meshes can offer more
(IST) IoT network. than one path when the path is longer than one ISL hop. Thus,
A. LEO Constellation-based IoT Network Architecture the requirements of routing strategies for DSTN are listed as
follows: 1) avoiding choosing multiple routes that will lead to
Combining illustrations in Section III and viewpoint of network, ISL congestion; 2) reducing the hops of route to increasing time
the LEO satellite design in a constellation is based on one of efficiency; 3) providing QoS guarantee. Considering the first
two fundamentally different approaches as follows [37]: two points, a kind of routing strategies is the variations of
1) Each satellite is a transparent retransmitter to relay traffic Dijkstra algorithm, which utilize the topology snapshot method
received from IoT terminals and earth gateway station [40]. Topology snapshot is to divide continual dynamic
(EGS), and returning the traffic to the ground. topology of satellite constellation network into a series of static
2) Each satellite with ISLs is a network switch for being able topology structure called “snapshot”, which update while the
to communicate with neighboring satellites. In this ISLs changing. Therefore, at every arbitrary moment, the
approach, IoT terminals within a satellite visible area DSTN has a particular snapshot that remains stable until next
(called satellite’s footprint) can exchange traffic with EGS changing appears, and routing strategies used frequently are
minimum distance algorithm (MDA) and minimum hops

2169-3536 (c) 2017 IEEE. Translations and content mining are permitted for academic research only. Personal use is also permitted, but republication/redistribution requires IEEE permission. See
http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2017.2735988, IEEE Access

algorithm (MHA) focusing on different ISL factors. For MDA,

UE eNodeB MME SGW
cost function of each ISL is defined as its physical distance,
while cost function is defined as one for MHA. Clearly, MDA
Performs access barring
is to find out the route with minimum distance that aims to check based on the
reduce the time delay of communication but ignore the whole information in MIB-NB
system flux balance, which may lead to severe ISL congestion. and SIB14-NB

Correspondingly, MHA focuses on decreasing route occupancy RA msg1

Preamble
on system resources but does not consider handover cost. In RA msg2
[40], an application of K-shortest path algorithm is proposed to Preamble
RA msg3
avoid ISL congestion and reduce handover cost at the same RRC Connection Resume Request

time. RRC Connection Resume

In [41], [42], routing strategies focusing on QoS guarantees RRC Connection Resume
are proposed. Unlike Dijkstra algorithm based routing Complete S1 UE Context Resume
Request
strategies that pre-determine routes in EGS, in these ones, each S1 UE Context Resume
satellite updates the local routing table according to its UL DATA
Response

signaling without the assistance of the terrestrial gateway UL DATA

stations, which improve the viability of DSTN. Satellites Modify Bearer Request

forecasts the routing signaling earlier before disconnection of Modify Bearer Response

the ISL and broadcast the signaling once the ISL is connected. UL/DL data packets

After receiving the routing signaling, each satellite updates the Fig. 8. Data transmission using RRC resume operation.
routing table according to the path state collected by its
signaling without any acknowledgement. Each intermediate UE
LEO EGS
MME SGW
satellite (DPG)
satellite fills the link state information between itself and its
previous satellite in the signaling record, which is referred to as Triggered into
“reverse detection”. When congestion occurred in one satellite, transmitting mode

it will generate a warning signal and send it to next satellite. RA msg1

Connection Establish Request
Then the destination satellite broadcast a detection request for (UE ID, key)

searching a new route with low congestion level. Moreover, Connection Establish
DL UE data packets

[22] proposes QoS mapping techniques over heterogeneous Connection Establish Complete Match UE context
UE data packets information
networks and an applicable solution called technological ACK
S1 UE Context Resume
independent-service access point (TI-SAP), which are reliable Turn into deep Request

solutions for LEO constellation-based IoT network. sleep mode S1 UE Context Resume
Response

B. United Higher Layer Design for IST IoT network UL DATA

Modify Bearer Request

Currently, in the field of LPWAN, there are several popular Modify Bearer Response
technologies such as NB-IoT, LoRa, Sigfox, Random Phase UL/DL data packets
Multiple Access (RPMA), etc. However, with respect to higher
layer design, NB-IoT has incomparable advantages over others
since it bases on LTE, which has a complete higher layer Fig. 9. Data transmission operation in LEO constellation-based IoT.
structure. In 3GPP Rel-13 [43], NB-IoT has been introduced to
provide LPWA IoT services by modifying LTE technology. request to mobility management entity (MME), there are
Thus, the higher layer design for IST IoT network may refer to five signaling interactions between UE and eNodeB.
the structure of NB-IoT. Unlike NB-IoT, in LEO constellation-based IoT, the
In NB-IoT, LTE functionalities are inherited with functionality of eNodeB is shared by satellite and EGS. As
simplifications and optimizations that only essential features mentioned in Section V-A, the connection between
for small data transmission are supported in Rel-13 [44]. A satellite UE and eNodeB is not stable physical connection
typical optimization for small data transmission is Radio but routing through a dynamic topology network, which
Resource Control (RRC) connection suspend/resume. A means each signaling interaction may transmit via
diagram of RRC resume operation is shown in Fig. 8. Both UE different route. Due to the limited resource of ISLs and
and eNodeB store the Access Stratum (AS) context together retransmitters, continually signaling interaction is bound
with Resume ID upon connection suspension. When UE turns to affect the system throughput and efficiency.
into data transmission progress, it provides the stored Resume 2) Context information storage: If utilize NB-IoT’s pattern,
ID to be used by eNodeB to access the stored information UE and satellite will store AS context and Resume ID
required to resume the RRC connection. upon connection suspension. However, the limited
However, several disadvantages can be seen from this RRC on-board memory cannot afford such cost.
resume operation that NB-IoT’s higher layer design cannot To release space segment pressure, proxy cache scheme can
fully meet the LEO constellation-based IoT system’s be also implemented in the LEO constellation-based IoT’s
requirements: higher layer design. The concrete measure is to separate
1) Continually signaling interaction: In RRC resume UE-satellite segment as a pure transmission network from the
operation, before eNodeB sending UE context resume terrestrial segment, e.g. the EGS. Data package oriented from

2169-3536 (c) 2017 IEEE. Translations and content mining are permitted for academic research only. Personal use is also permitted, but republication/redistribution requires IEEE permission. See
http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2017.2735988, IEEE Access

UE contain the UE ID and corresponding key and satellite play TABLE III
the role of authentication. The UE-satellite segment is similar INTERFERENCE ANALYSIS PARAMETERS
to the transmitting part of LoRa [45]. Upon finishing
transmission, UE turns into deep sleep mode until next Satellite altitude 780 km
transmitting trigger. The data processing gateway (DPG) in Satellite uplink TX EIRP 33 dBm -38.5 dBm
EGS, which utilizes proxy cache scheme, stores all context Satellite RX antenna gain 17dBi (noise figure 8dB)
information for establishing and resuming connection. After RX G/T -5dB/K
receiving the satellite downlink data, DPG will identifies Interference source TX
different UE’s data packages and then simulates corresponding 0 dBm
power
UE to complete higher layer operation using stored information. Abbreviations
The diagram is depicted in Fig. 9. By implementing DPG, space EIRP equivalent isotopically radiated power
segment pressure can be sharply released. TX transmitter
RX receiver

VI. LEO CONSTELLATION-BASED IOT INTERFERENCE

ANALYSIS TABLE IV
INTERFERENCE ANALYSIS OF ELEVATION ANGLE
Commonly, current satellite operating frequency tends to
deploy in higher frequency bands such as Ku and Ka band, Maximum
which can improve system capacity. However, operating on Bearable TISs
TX EIRP (dBm) Elevation Angle (BER )
those bands cannot meet the requirements of IoT UE (e.g. small (degree)
size and low power consumption) due to the severe propagation
33 47 4,320,000
impairments including path loss and rain attenuation [46].
36 60 3,376,000
Meanwhile, dedicated lower frequency bands for mobile
38.5 72 2,900,000
satellite system are rare that only a few companies, e.g.
Orbcomm and Iridium, have registered for licensed radio
spectrum. Despite operating on licensed frequencies, those TABLE V
systems need to negotiate with local radio administration since SRRC TABLE OF ALLOCATIONS IN LORA ISM BANDS
there is not a global pattern for frequency allocation. Therefore,
sharing the same frequency bands with terrestrial IoT systems Frequency bands China Mainland
spontaneously becomes a considerable option for LEO Broadcasting
constellation-based IoT. There are two existed frequency plans Space operation (space-earth)
corresponding to LoRa and NB-IoT, respectively. LoRa 470-510 MHz Space research (space-earth)
alliance [45] suggests operating on unlicensed Industrial [Fixed services]
Scientific Medical (ISM) band referring to local radio standard. [mobile services]
On the contrary, NB-IoT operates on licensed radio spectrum Broadcasting
779-787 MHz [Fixed services]
with three different operation modes: (1) stand-alone as a
[mobile services]
dedicated carrier, (2) in-band within the occupied bandwidth of
a wideband LTE carrier, and (3) within the guard-band of an
existing LTE carrier [10]. where is the link distance in meters and is the carrier
Unfortunately, sharing the same frequency band between frequency in Hertz. The signal-to-interference ratio (SIR) can
terrestrial and satellite networks may lead interference to both be expressed in decibels by
satellite and terrestrial cells [47]. Usually, a LEO satellite spot (5)
beam diameter can be over a thousand kilometers, which can
where the calculation of is as follows:
accumulate from a large area including highly populated cities
crowed by terrestrial cells. Even though using multi-beam (6)
antenna and designing LEO and terrestrial networks to cover
where is the number of terrestrial interference sources
different areas, the satellite beams can drift and cause
overlapping in the satellite and terrestrial cells. (TISs). To be noticed that The TISs increase with the decrease
In [47], an analysis on satellite uplink transmission with of elevation angle due to the increase of satellite spot center
terrestrial network interference has been proposed. The distance. Meanwhile, it assumes that the interference signals
simulated parameters are listed in Table III. The received signal are added together and then multiplied with
power at satellite antenna can be calculated as considering the numerous interference sources, where is
(3) defined by
where is the transmitter power in dBm, is the gain of (7)
satellite antenna, and is propagation loss calculated by City Toulouse, France and Dusseldorf, Germany are taken
(4) into analysis in [47]. Though the number of TISs in Toulouse at
the elevation angle of is 435,180 and is 10 times larger
than that at the elevation angle of , the former one’s SIR

2169-3536 (c) 2017 IEEE. Translations and content mining are permitted for academic research only. Personal use is also permitted, but republication/redistribution requires IEEE permission. See
http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2017.2735988, IEEE Access

is 20dB higher than the later one’s, which is because the not allowed to access the medium concurrently [51],
simultaneously increasing propagation loss counteract the including spectral energy, cyclostationary detection and
accumulated interference. Therefore, performance of a SIR estimation [52], and shares information with
satellite-terrestrial hybrid system depending on population databases to update the real-time interference scenario.
density and elevation angle. Analysis results for Dusseldorf The core function of SE unit is dynamic channel assignment
referring to elevation angle and TISs are shown in Table IV. (DCA), which directly has respect to interference mitigation.
The analysis proves the potential feasibility for a Synthesizing the cognitive information collected by SA unit,
satellite-terrestrial hybrid system sharing the same frequency DCA is necessary to allocate the determined cognitive
band. However, the results show that satellite link is not resources, particularly the frequency channels, to UE. [49]
immune to terrestrial interference especially for areas with high proposes two allocating approaches focus on maximizing
population density. Meanwhile, several assumptions in [47] are overall system throughput and system availability, respectively.
not available considering the LEO constellation-based IoT In the allocating process, to ensure efficiency, factors such as
system; for instance, uplink TX ERIP is too high to meet the channel priorities and aggregation need to be considered. With
UE requirement of low power consumption. Therefore, the respect to channel priorities, DCA is designed to evaluate the
hybrid system cannot be designed solely on spectrum sharing, channel interference level in an adjustable time interval.
and further anti-interference measures should be taken into Suitable channels are prioritized in terms of interference power
consideration. expected on the next scan, which bases on SIR estimation
Two topics are worth considering in the framework of implemented in spectrum sensing. Meanwhile, SE unit also
potential anti-interference measures: 1) technology of cognitive performs power control in the downlink, which keep
radios (CRs) dealing with interference mitigation in satellite interference to the incumbent users below a specific threshold
networks; 2) spread spectrum (SS) techniques, namely direct on the premise of maintaining the system throughput.
sequence (DS) and frequency-hopping spread spectrum (FHSS), Apart from aforementioned techniques, several methods
in satellite communication (SATCOM). These topics are including compressive sensing (CS) and beamforming are used
discussed below. in CRs mechanisms. In [53], CS is suggested to operate channel
estimation due to the sparsity inherent in wireless channels.
A. Cognitive Radios Mechanisms for Interference mitigation
Since the active transmitting UE cannot be predicted in advance,
CRs, from its definition, is a radio system that is able to sense the sparsity condition in CS theory can be justified.
its operating electromagnetic environment and adjust the radio Beamforming, as an essential technique in 5G, can be
parameters to optimize system performance dynamically and implemented on satellite for interference detection and
automatically [48]. In order to utilize shared frequency bands, improving SIR of UEs [54].
CRs is needed while relieving interference from the terrestrial
B. SS Techniques in SATCOM
services and guaranteeing acceptable interference to the
incumbent users [49]. SS technique is a common method implemented in wireless
With special focus on Chinese Mainland, LoRa can be used communication systems to resist interference from same
in 470-510 MHz band and 779-787 MHz band in line with frequency band. In current IoT standards, LoRa adopt
LoRa alliance regulations [50] and State Radio Regulation of DS/CDMA for its physical layer transmitting protocol due to
China (SRRC). Table V provides the SRRC allocations of the the unlicensed frequency plan. Therefore, DS/CDMA can be
aforementioned frequency bands, where the services expressed also considered for LEO constellation-based IoT. In [55], total
in bracket are less important. As can be seen from Table IV, capacity in a shared CDMA/LEO environment is calculated,
LEO constellation-based IoT deploying in LoRa frequency which presents a theoretical base for implementing DS/CDMA
plan is sharing frequency bands with other radio services, in LEO constellation-based IoT. Furthermore, to improve
which means in downlink, UEs need to manage interference system efficiency, an adjustable spreading factor (SF) scheme
received from the incumbent users, and in uplink, UEs need to is recommended by LoRa alliance [56]. Under the circumstance
ensure their transmission does not impact the incumbent of LEO constellation-based IoT, basing on the channel state
receivers. Therefore, CRs mechanisms is necessary to mitigate information broadcasted by satellite, UE can adaptively adjust
the interference issues. SF and data rate in order to increase system throughput and
A spectrum utilization system with CRs mechanisms maintain UE power efficiency.
consists spectrum awareness (SA) unit and spectrum FH is another approach of SS and extensively used in
exploitation (SE) unit. The SA unit aims to gather knowledge military satellite systems to resist intentional and unintentional
about the incumbent users with the cooperation between interference. A fixed pseudorandom noise (PN) sequence is
databases and spectrum sensing. used in traditional FH operation, known to both transmitter and
1) Databases: The purpose of databases is to incorporate receiver, to switch a carrier among available channels. The
satellite broadcasting links and terrestrial services limitation of traditional FH, however, is that PN sequence will
characteristics in order to determine terrestrial interference inevitably switch a carrier that is persistently occupied by other
levels at any location and carrier frequency. Once received users since both transmitter and receiver have no channel state
interference level is deemed to exceed threshold, CRs information. To solve the limitation of static FH scheme, [57]
mechanisms then ought to mitigate the interference. proposes a dynamic frequency hopping (DFH) scheme
2) Spectrum Sensing: Spectrum sensing provides the combining with CRs techniques. In this approach, FH sequence
essential information to enable this interweave is no longer a fixed PN sequence but depending on the spectrum
communications in which primary and secondary users are sensing information. Basing on proactive sensing, DFH could

2169-3536 (c) 2017 IEEE. Translations and content mining are permitted for academic research only. Personal use is also permitted, but republication/redistribution requires IEEE permission. See
http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2017.2735988, IEEE Access

mitigate both intended and unintended interference. International Conference on Telecommunications and Multimedia
(TEMU), 2014, pp. 132-137.
[14] ORBCOMM. Industrial M2M tracking, Monitoring and Control
VII. CONCLUSIONS Solutions. [Online]. Available: http://www.orbcomm.com
[15] Y. Kawamoto, H. Nishiyama, N. Kato, S. Yamamoto, N. Yoshimura, and
This paper presents an overview of LEO satellite
N. Kadowaki, "On real-time data gathering in next generation
constellation-based IoT system. Important issues that have satellite-routed sensor system (SRSS)," in 2012 International Conference
been discussed are: constellation design, interference on Wireless Communications and Signal Processing (WCSP), 2012, pp.
mitigation, constellation network architecture, routing scheme, 1-5.
and united higher layer design. For each of these topics, the [16] Y. Kawamoto, H. Nishiyama, Z. M. Fadlullah, and N. Kato, "Effective
Data Collection Via Satellite-Routed Sensor System (SRSS) to Realize
authors report results of recent studies and post some specific Global-Scaled Internet of Things," IEEE Sensors Journal, vol. 13, pp.
thinking for system design. 3645-3654, 2013.
As mentioned in [6], potential use of constellations of [17] Y. E. Ning, Y. Zhu, R. C. Wang, R. Malekian, and Q. M. Lin, "An
satellites for IoT applications is of growing interests. With Efficient Authentication and Access Control Scheme for Perception
Layer of Internet of Things," Applied Mathematics & Information
booming development in IoT environment, as a powerful Sciences, vol. 8, pp. 1617-1624, 2014.
supplement to terrestrial systems, LEO constellation-based IoT [18] P. K. Gupta, B. T. Maharaj, and R. Malekian, "A novel and secure IoT
is worth being focused and studied. To make this topic become based cloud centric architecture to perform predictive analysis of users
activities in sustainable health centres," Multimedia Tools & Applications,
a reliable cost-benefit solution, further researches are needed to
pp. 1-24, 2016.
be done including transmission scheme, system security and [19] A. C. Jose, R. Malekian, and N. Ye, "Improving Home Automation
low power consumption design. Security; Integrating Device Fingerprinting Into Smart Home," IEEE
Access, vol. 4, pp. 5776-5787, 2016.
[20] A. C. Jose and R. Malekian, "Improving Smart Home Security:
ACKNOWLEDGMENT Integrating Logical Sensing Into Smart Home," IEEE Sensors Journal,
This work is supported by the National Natural Science vol. 17, pp. 4269-4286, 2017.
Foundation of China (Grant No.91338201). [21] A. Detti, A. Caponi, and N. Blefari-Melazzi, "Exploitation of Information
Centric Networking principles in satellite networks," in 2012 IEEE First
AESS European Conference on Satellite Telecommunications (ESTEL),
REFERENCES 2012, pp. 1-6.
[22] M. Marchese, QoS Over Heterogeneous Networks: Wiley Publishing,
2007.
[23] H. Bedon, C. Miguel, A. Fernandez, and J. S. Park, "A DTN System for
[1] L. Atzori, A. Iera, and G. Morabito, "The Internet of Things: A survey,"
Nanosatellite-based Sensor Networks using a New ALOHA Multiple
Computer Networks, vol. 54, pp. 2787-2805, 2010.
Access with Gateway Priority," Smartcr, vol. 3, 2013.
[2] A. Zanella, N. Bui, A. Castellani, L. Vangelista, and M. Zorzi, "Internet
[24] S. Burleigh, V. Cerf, R. Durst, K. Fall, A. Hooke, K. Scott, et al., "The
of Things for Smart Cities," IEEE Internet of Things Journal, vol. 1, pp.
interplanetary internet: A communications infrastructure for Mars
22-32, 2014.
[3] P. Bellavista, G. Cardone, A. Corradi, and L. Foschini, "Convergence of exploration ☆," Acta Astronautica, vol. 53, pp. 365-373, 2003.
MANET and WSN in IoT Urban Scenarios," IEEE Sensors Journal, vol. [25] Q. Xu, H. Zhang, Y. Cheng, S. Zhang, and W. Zhang, "Monitoring and
13, pp. 3558-3567, 2013. Tracking the Green Tide in the Yellow Sea With Satellite Imagery and
[4] MachinaResearch. (2016, Aug). Press Release: Global Internet of Things Trajectory Model," IEEE Journal of Selected Topics in Applied Earth
Market to Grow to 27 Billion Devices, Generating USD 3 Trillion Observations and Remote Sensing, vol. 9, pp. 5172-5181, 2016.
Revenue in 2025 [Online]. Available: [26] M. Faruolo, I. Coviello, T. Lacava, N. Pergola, and V. Tramutoli, "A
https://machinaresearch.com/news/press-release-global-internet-of-thing Multi-Sensor Exportable Approach for Automatic Flooded Areas
s-market-to-grow-to-27-billion-devices-generating-usd3-trillion-revenue Detection and Monitoring by a Composite Satellite Constellation," IEEE
-in-2025/ Transactions on Geoscience and Remote Sensing, vol. 51, pp. 2136-2149,
[5] G. Fairhurst, L. Caviglione, and B. Collini-Nocker, "FIRST: Future 2013.
Internet - a role for satellite technology," in 2008 IEEE International [27] M. Mityagina and O. Lavrova, "Satellite monitoring of the Black Sea
Workshop on Satellite and Space Communications, 2008, pp. 160-164. surface pollution," in 2015 IEEE International Geoscience and Remote
[6] M. D. Sanctis, E. Cianca, G. Araniti, I. Bisio, and R. Prasad, "Satellite Sensing Symposium (IGARSS), 2015, pp. 2291-2294.
Communications Supporting Internet of Remote Things," IEEE Internet [28] J. Marihart, "Communications technology guidelines for EMS/SCADA
of Things Journal, vol. 3, pp. 113-123, 2016. systems," IEEE Transactions on Power Delivery, vol. 16, pp. 181-188,
[7] Coverage Maps. Available: https://opensignal.com/networks 2001.
[8] "Satellite Machine-to-Machine (M2M) Communications - Global [29] S. M. Amin and B. F. Wollenberg, "Toward a smart grid: power delivery
Strategic Business Report," Global Industry Analysts, Inc, San Jose, CA, for the 21st century," IEEE Power & Energy Magazine, vol. 3, pp. 34-41,
USA, July 2016, 197 pp. 2005.
[9] R. Ratasuk, B. Vejlgaard, N. Mangalvedhe, and A. Ghosh, "NB-IoT [30] Q. Yang, D. I. Laurenson, and J. A. Barria, "On the Use of LEO Satellite
system for M2M communication," in 2016 IEEE Wireless Constellation for Active Network Management in Power Distribution
Communications and Networking Conference, 2016, pp. 1-5. Networks," IEEE Transactions on Smart Grid, vol. 3, pp. 1371-1381,
[10] N. Mangalvedhe, R. Ratasuk, and A. Ghosh, "NB-IoT deployment study 2012.
for low power wide area cellular IoT," in 2016 IEEE 27th Annual [31] L. Yushi, J. Fei, and Y. Hui, "Study on application modes of military
International Symposium on Personal, Indoor, and Mobile Radio Internet of Things (MIOT)," in 2012 IEEE International Conference on
Communications (PIMRC), 2016, pp. 1-6. Computer Science and Automation Engineering (CSAE), 2012, pp.
[11] A. Rico-Alvarino, M. Vajapeyam, H. Xu, X. Wang, Y. Blankenship, J. 630-634.
Bergman, et al., "An overview of 3GPP enhancements on machine to [32] A. Kott, A. Swami, and B. J. West, "The Internet of Battle Things,"
machine communications," IEEE Communications Magazine, vol. 54, pp. Computer, vol. 49, pp. 70-75, 2016.
14-21, 2016. [33] A. H. Ballard, "Rosette Constellations of Earth Satellites," IEEE
[12] G. Cocco and C. Ibars, "On the Feasibility of Satellite M2M Systems," in Transactions on Aerospace and Electronic Systems, vol. AES-16, pp.
Aiaa International Communications Satellite Systems Conference, 2013. 656-673, 1980.
[13] S. Vassaki, G. T. Pitsiladis, C. Kourogiorgas, M. Poulakis, A. D. [34] F. Bastia, C. Bersani, E. A. Candreva, S. Cioni, G. E. Corazza, M. Neri, et
Panagopoulos, G. Gardikis, et al., "Satellite-based sensor networks: al., "LTE Adaptation for Mobile Broadband Satellite Networks,"
M2M Sensor communications and connectivity analysis," in 2014 EURASIP Journal on Wireless Communications and Networking, vol.
2009, pp. 1-13, 2009.

2169-3536 (c) 2017 IEEE. Translations and content mining are permitted for academic research only. Personal use is also permitted, but republication/redistribution requires IEEE permission. See
http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2017.2735988, IEEE Access

[35] G. Liva, "Graph-Based Analysis and Optimization of Contention

Resolution Diversity Slotted ALOHA," IEEE Transactions on
Communications, vol. 59, pp. 477-487, 2011.
[36] E. Casini, R. D. Gaudenzi, and O. D. R. Herrero, "Contention Resolution
Diversity Slotted ALOHA (CRDSA): An Enhanced Random Access
Schemefor Satellite Access Packet Networks," IEEE Transactions on
Wireless Communications, vol. 6, pp. 1408-1419, 2007.
[37] L. Wood, G. Pavlou, and B. Evans, "Effects on TCP of routing strategies
in satellite constellations," IEEE Communications Magazine, vol. 39, pp.
172-181, 2001.
[38] G. R. Thomas, "The Manhattan Street Network," in Performance
Evaluation and High Speed Switching Fabrics and Networks:ATM,
Broadband ISDN, and MAN Technology, ed: Wiley-IEEE Press, 1993, pp.
303-309.
[39] N. Maxemchuk, "Routing in the Manhattan Street Network," IEEE
Transactions on Communications, vol. 35, pp. 503-512, 1987.
[40] J. f. He, Y. Jiang, D.-m. Bian, and G. x. Li, "Routing strategy research
based on ISL states and topology snapshot in LEO satellite constellation,"
in 2008 11th IEEE International Conference on Communication
Technology, 2008, pp. 13-16.
[41] J. Yan, Y. Zhang, and Z. Cao, "Reverse detection based QoS routing
algorithm for LEO satellite constellation networks," Tsinghua Science
and Technology, vol. 16, pp. 358-363, 2011.
[42] Y. Zhang, L. Zhu, J. Wu, H. Jiang, J. Zhou, T. Cao, et al., "A Novel QoS
Routing Method for LEO Rosette Constellation Network," in 2015 IEEE
International Conference on Computer and Information Technology;
Ubiquitous Computing and Communications; Dependable, Autonomic
and Secure Computing; Pervasive Intelligence and Computing, 2015, pp.
525-531.
[43] 3GPP. (2015). Release 13 [Online]. Available:
http://www.3gpp.org/release-13
[44] R. Ratasuk, N. Mangalvedhe, Y. Zhang, M. Robert, and J. P. Koskinen,
"Overview of narrowband IoT in LTE Rel-13," in 2016 IEEE Conference
on Standards for Communications and Networking (CSCN), 2016, pp.
1-7.
[45] Lora alliance [Online]. Available: http://www.lora-alliance.org/
[46] A. D. Panagopoulos, P.-D. M. Arapoglou, and P. G. Cottis, "Satellite
Communications At Ku Ka And V Bands:Propagation Impairments and
Mitigation Techniques," IEEE Communications Surveys, vol. 6, pp. 2-14,
2004.
[47] J. Janhunen, J. Ketonen, A. Hulkkonen, J. Ylitalo, A. Roivainen, and M.
Juntti, "Satellite Uplink Transmission with Terrestrial Network
Interference," in 2015 IEEE Global Communications Conference
(GLOBECOM), 2015, pp. 1-6.
[48] T. Yucek and H. Arslan, "A survey of spectrum sensing algorithms for
cognitive radio applications," IEEE Communications Surveys &
Tutorials, vol. 11, pp. 116-130, 2009.
[49] S. Maleki, S. Chatzinotas, B. Evans, K. Liolis, J. Grotz, A.
Vanelli-Coralli, et al., "Cognitive spectrum utilization in Ka band
multibeam satellite communications," IEEE Communications Magazine,
vol. 53, pp. 24-29, 2015.
[50] "LoRaWAN Regional Parameters," ed: LoRa alliance, 2016.
[51] A. Ali and W. Hamouda, "Advances on Spectrum Sensing for Cognitive
Radio Networks: Theory and Applications," IEEE Communications
Surveys & Tutorials, vol. 19, pp. 1277-1304, 2017.
[52] E. Axell, G. Leus, E. G. Larsson, and H. V. Poor, "Spectrum Sensing for
Cognitive Radio : State-of-the-Art and Recent Advances," IEEE Signal
Processing Magazine, vol. 29, pp. 101-116, 2012.
[53] M. Jia, X. Gu, Q. Guo, W. Xiang, and N. Zhang, "Broadband Hybrid
Satellite-Terrestrial Communication Systems Based on Cognitive Radio
toward 5G," IEEE Wireless Communications, vol. 23, pp. 96-106, 2016.
[54] S. K. Sharma, S. Maleki, S. Chatzinotas, J. Grotz, J. Krause, and B.
Ottersten, "Joint Carrier Allocation and Beamforming for cognitive
SatComs in Ka-band (17.3–18.1 GHz)," in 2015 IEEE
International Conference on Communications (ICC), 2015, pp. 873-878.
[55] B. R. Vojcic, L. B. Milstein, and R. L. Pickholtz, "Total capacity in a
shared CDMA LEOS environment," IEEE Journal on Selected Areas in
Communications, vol. 13, pp. 232-244, 1995.
[56] "LoRaWAN Specification," ed: LoRa Alliance, 2016.
[57] I. Aykin and M. Krunz, "Proactive Sensing and Interference Mitigation in
Multi-Link Satellite Networks," in 2016 IEEE Global Communications
Conference (GLOBECOM), 2016, pp. 1-6.

2169-3536 (c) 2017 IEEE. Translations and content mining are permitted for academic research only. Personal use is also permitted, but republication/redistribution requires IEEE permission. See
http://www.ieee.org/publications_standards/publications/rights/index.html for more information.