Upload
khangminh22
View
0
Download
0
Embed Size (px)
Citation preview
For Peer Review
1
Evolution and Future Trends of Research in
Cognitive Radio: A Contemporary Survey
Ekram Hossain, Dusit Niyato, and Dong In Kim
Abstract
The cognitive radio (CR) paradigm for designing next generation wireless communications systems is becoming
increasingly popular and different aspects of it are being implemented in currently available wireless systems.
In the last decade, significant amount of research efforts have been made to solve CR challenges and several
standards related to cognitive radio and dynamic spectrum access have been developed. Also, there have been
advances in software-defined radio (SDR) platforms to implement the cognitive radio systems. In this article,
we provide a comprehensive survey on the evolution of CR research covering aspects such as spectrum sensing,
measurements and statistical modeling of spectrum usage, physical layer (PHY) aspects such as waveform and
modulation design, multiple access, resource allocation and power control, cognitive learning, adaptation and self-
configuration, multi-hop transmission and routing, and robustness and security in cognitive radio networks. Also,
state-of-the-art research on the economics of cognitive radio networks, cognitive radio simulation tools, testbeds
and hardware prototypes, cognitive radio applications, and cognitive radio standardization efforts are summarized.
Emerging trends on cognitive radio research and open research challenges related to the cost-effective and large-
scale deployment of cognitive radio systems are outlined.
I. INTRODUCTION AND HISTORICAL NOTES ON COGNITIVE RADIO
For wireless communications, the right to access the spectrum or license is generally defined by
frequency, space, transmission power, spectrum owner (i.e., licensee), type of use, and the duration of
license. Normally, a license is assigned to one licensee, and the use of spectrum by this licensee must
conform to the specification in the license (e.g., maximum transmission power and type of use). In the
traditional spectrum licensing scheme, the license cannot change the type of use or transfer the right to
other licensees. Also, the radio spectrum is licensed for a large region and generally in large chunks.
All these factors in the current command-and-control model for spectrum allocation and assignment limit
the use and result in low utilization of the frequency spectrum. Since the existing and the new wireless
Page 3 of 66
http://mc.manuscriptcentral.com/wcm
Wireless Communications and Mobile Computing
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
For Peer Review
2
applications and services are demanding for more transmission capacity and data transmission performance
and hence more radio spectrum, the utilization of the radio spectrum needs to be improved.
To improve the efficiency and utilization of the radio spectrum, the limitations above should be remedied
by modifying the spectrum licensing scheme and adopting a dynamic spectrum management model. The
idea is to make spectrum access more flexible by allowing the unlicensed users to access the radio
spectrum under certain restrictions. Since the legacy wireless systems were designed to operate on a
dedicated frequency band, they are not able to utilize the improved flexibility provided by this spectrum
licensing scheme. Therefore, the concept of cognitive radio emerged the main goal of which is to provide
adaptability to wireless transmission through dynamic spectrum access so that the utilization of the
frequency spectrum can be enhanced without losing the benefits associated with static spectrum allocation.
The cognitive radio is a “smarter radio” in the sense that it can sense channels that contain signals from
a large class of heterogeneous devices, networks, and services. Based on this sensing, the radio will
implement sophisticated algorithms to share the limited-bandwidth channel with other users in order to
achieve efficient wireless communication. In this way, the cognitive radio concept generalizes the idea of
multiple access involving devices in a single homogeneous system to multiple access among devices in
different radio spectrum using different radio transmission techniques and hence different systems (i.e.,
inter-system multiple access as opposed to the more traditional intra-system multiple access) which have
different priorities in accessing the spectrum [1].
Again, the concept of “cognitive radio” can be generalized as a topic within the field of adaptive or
cognitive systems. The cognitive radio refers to a sophisticated radio device that mimics the human brain
in a dynamic spectrum access (DSA) environment to perceive and learn the radio environment to control
and adapt the transmission actions [2]. Such a radio device or terminal has to operate in a networked
environment where many of such devices sense the radio channels in their own geographical environment
and the sensed information is processed (either centrally or distributively) to make a decision on channel
access using appropriate modulation scheme, transmission power and error control algorithms. The wireless
network infrastructure will also have a significant role here, for example, to manage information about
the radio devices and their operating environments, determine the consequences of interference, facilitate
cooperation among the radio devices, etc. Significant challenges exist in the design, analysis, optimization,
development, and deployment of DSA-based cognitive radio networks (CRNs), and these challenges in-
volve both technology and policy (or regulatory) considerations. The major technical challenges include the
following: spectrum sensing, measurements and statistical modeling of radio spectrum usage; waveform,
Page 4 of 66
http://mc.manuscriptcentral.com/wcm
Wireless Communications and Mobile Computing
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
For Peer Review
3
modulation, and RF front-end design at the physical layer; methods for multiple access, resource allocation
and transmission power control for cognitive radios; distributed learning, adaptation, self-configuration,
cooperation, and coexistence methods for cognitive radios; routing in multi-hop CRNs and cross-layer
optimization; methods for robustness and security in CRNs; wide-area end-to-end performance modeling
and analysis for global internetworking; modeling and understanding the emergent system behavior;
development of network control and management protocols; software abstractions for the radio modules at
the application programming interface (API) level; and development of testbeds and real-world deployment
of CRNs. The challenges in the spectrum policy domain include the following: development of practical
DSA policies that lead to efficient spectrum use, protect the rights of license holders, and maintain the
service quality; development of certification and enforcement policies for cognitive radios [3].
The term “cognitive radio” was first used and discussed in [4] in 1999 by J. Mitola III and G. Q.
Maguire. In 2000, in his Ph.D. dissertation [5], J. Mitola III introduced cognitive radio and networks as
an extension of the SDR concept where the radios are computationally intelligent and are aware of the
radio environment and user applications. Although the term “cognitive radio” has been used only recently,
the concept of intelligent spectrum access is not completely new. For example, the idea of sensing before
transmitting as used in the popular carrier-sense multiple access (CSMA) protocols dates back to early
90’s when the IEEE 802.3 standards or Ethernet standards were being developed. Also, the idea of using
“dynamic channel allocation and selection” was extensively studied in the 80’s and 90’s in the context of
mobile cellular wireless networks. The fact that one of the main reasons of the spectrum scarcity problem
is the traditional fixed spectrum allocation, was first realized when the FCCs Spectrum Policy Task Force
Report [6] was published in 2002. Since then, FCC pushed towards advances in technology for more
efficient spectrum usage, and the idea of exploiting cognitive radio technology to facilitate the flexible,
efficient, and reliable use of spectrum gained momentum in both industry and academia. In particular, FCC
advocated the use of the concepts such as dynamic frequency selection (DFS), incumbent profile detection
(IPD), and transmit power control (TPC) for cognitive radios. In 2004, FCC issued an NPRM (Notice of
Proposed Rulemaking) which introduced the possibility of allowing the unlicensed (or secondary) users
to temporarily use spectrum allocated to licensed (or primary) users as long as the secondary users do
not cause harmful interference to primary users. An important milestone in the development of cognitive
radio is the approval by FCC in November 2008 of the unlicensed use of the TV white space (54-72 MHz,
76-88 MHz, 174-216 MHz, and 470-806 MHz) based on spectrum sensing as well as consultation with an
FCC-mandated database [7]. These spectrum bands are attractive due to their superior radio propagation
Page 5 of 66
http://mc.manuscriptcentral.com/wcm
Wireless Communications and Mobile Computing
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
For Peer Review
4
characteristics in the sub-gigahertz frequencies. In September 2010, FCC released new rules for the use
of white space for unlicensed wireless devices which removed the mandatory sensing requirements and
thus facilitated the use of the spectrum with geolocation based channel allocation [8]. This has motivated
the development of new wireless standards and standardization initiatives including the IEEE 802.22 for
wireless regional area networks (WRANs), the European Computer Manufacturers Association (ECMA)
392 for personal portable devices, the IEEE 802.11af and 802.19.1 Task Groups.
In the last decade, there have been a significant amount of research work focusing on different aspects
in CRNs such as spectrum sensing and signal classification techniques at the physical layer, scheduling,
transmission power control, and adaptive channel access protocols at the link layer, resource allocation
for multi-hop transmission at the routing layer, modeling and analysis of security, and cognitive radio
network economics. In addition to a huge number of research papers including survey and tutorial type of
articles [2], [4], [9]- [14], a number of research monographs and edited books [15]- [23] have appeared
in the literature. A number of special issues on cognitive radio have been organized in both IEEE and
non-IEEE journals [24]- [36]. A number of small-scale testbeds have been built in the academia. Also,
significant amount of efforts have been put by the different standardization bodies for the standardization
of CRNs. The aim of this article is to concisely summarize the current state-of-the-art research in the
dynamic spectrum access-based cognitive radio systems and the future research trends. The rest of the
article is organized as follows. After reviewing the fundamental concepts of cognitive radio and dynamic
spectrum access, we describe the building blocks of a CRN and outline the major research issues from
a layered perspective. Then we provide a summary of the current state-of-the-art spectrum sensing and
spectrum sharing methods for cognitive radio systems. This is followed by a concise review of the work on
robustness and security in CRNs, and on the economics of CRNs. Then we describe the available cognitive
radio simulation tools, testbeds, and hardware prototypes. The current and future trends in cognitive
radio including the applications (e.g., smart grid [37] [38], machine-to-machine communications [39],
and cloud computing [40]) are reviewed and the open research issues are outlined next. To this end, the
standardization activities on cognitive radio are summarized.
II. FUNDAMENTAL PRINCIPLES OF COGNITION AND DYNAMIC SPECTRUM ACCESS
The term “cognitive radio” defined in [2] is as follows: Cognitive radio is an intelligent wireless
communication system that is aware of its ambient environment. A cognitive radio transmitter will learn
from the environment and adapt its internal states to statistical variations in the existing RF stimuli by
Page 6 of 66
http://mc.manuscriptcentral.com/wcm
Wireless Communications and Mobile Computing
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
For Peer Review
5
adjusting the transmission parameters (e.g., frequency band, modulation mode, and transmission power)
in real-time and on-line manner. This definition essentially captures the fundamental concept behind
cognitive radio. A cognitive radio network enables us to establish communications among cognitive
radio nodes/users. The communication parameters can be adjusted according to the change in the radio
environment, topology, operating conditions, or user requirements. Two main objectives of the cognitive
radio are to improve the utilization of the frequency spectrum and to achieve the highly reliable and highly
efficient wireless communications.
Cognitive radio-based dynamic spectrum access or sharing has basically two major flavors, i.e., hor-
izontal spectrum sharing and vertical spectrum sharing. In the former case, all users/nodes have equal
regulatory status while in the latter case all users/nodes do not have equal regulatory status. In vertical
spectrum sharing, there are primary (i.e., licensed) users and secondary (i.e., unlicensed) users, and the
secondary users opportunistically access the spectrum without affecting the primary users’ performances.
Horizontal spectrum sharing can be between homogeneous networks (e.g., IEEE 802.11a operating in the
5 GHz Unlicensed National Information Infrastructure [U-NII] band) or between heterogeneous networks
(e.g., coexistence between IEEE 802.11b and 802.15.1 [Bluetooth] networks). When all the networks in
a heterogeneous environment have cognitive/adaptive capabilities (i.e., all coexisting networks have equal
incentives to adapt), it is referred to as symmetric sharing. On the other hand, when there is one or more
network without cognitive/adaptive capabilities (e.g., coexistence of legacy technology with cognitive radio
technology), this is referred to as asymmetric spectrum sharing [15]. One example of this is the coexistence
of high-speed IEEE 802.11 networks with low-power IEEE 802.15.4 networks. Dynamic spectrum access
in vertical spectrum sharing is referred to as opportunistic spectrum access. This opportunistic spectrum
access is the method for the secondary user to operate within a frequency band which is designated to
the primary user.
The major functionalities of a cognitive radio device include spectrum sensing, spectrum management,
and spectrum mobility [22]. Through spectrum sensing, the information of the target radio spectrum (e.g.,
type and current activity of licensed user) has to be obtained so that this information can be utilized
by the cognitive radio user. The spectrum sensing information is exploited by the spectrum management
function to analyze the spectrum opportunities and make decisions on spectrum access. If the status of
the target spectrum changes, the spectrum mobility function can switch the operating frequency bands for
the cognitive radio users.
• Spectrum sensing: The goal of spectrum sensing mechanism is to determine the status of the spectrum
Page 7 of 66
http://mc.manuscriptcentral.com/wcm
Wireless Communications and Mobile Computing
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
For Peer Review
6
(e.g., to detect the signature of a signal from a licensed user) and the activity of licensed user by
periodically sensing the target frequency band. In particular, a cognitive radio transceiver detects an
unused spectrum or spectrum hole (i.e., band, location, and time) and also determines the method of
accessing it (i.e., transmission power and access duration) without interfering the transmission of a
licensed user.
Spectrum sensing can be either centralized or distributed. In centralized spectrum sensing, a sensing
controller (e.g., access point or base station) senses the target frequency band, and the information
obtained from sensing is shared with other nodes in the system. Centralized spectrum sensing can
reduce the complexity of user terminals, since all the sensing functions are performed at the sensing
controller. However, centralized spectrum sensing suffers from location diversity. For example, the
sensing controller may not be able to detect an unlicensed user at the edge of the cell. In distributed
spectrum sharing, unlicensed users perform spectrum sensing independently, and the spectrum sensing
results can be either used by individual cognitive radios (i.e., non-cooperative sensing) or can be
shared among other users (i.e., cooperative sensing). Although cooperative sensing incurs more
communication and processing overhead, the accuracy of spectrum sensing is higher than that of
non-cooperative sensing.
• Spectrum analysis: The information obtained from spectrum sensing is used to schedule and plan
the spectrum access by the unlicensed users. In this case, the communication requirements of the
unlicensed users are also used to optimize the transmission parameters. Major components in spec-
trum management mechanisms are spectrum analysis and spectrum access optimization. In spectrum
analysis, information from spectrum sensing is analyzed to understand the ambient RF environment
(e.g., the behavior of licensed users) and gain knowledge about the spectrum holes (e.g., interference
estimation, duration of availability, and probability of collision with licensed user due to sensing
error). A knowledge base of the spectrum access environment can be built and maintained based on
learning and knowledge extraction. Machine learning algorithms from the field of artificial intelligence
can be applied for learning and knowledge extraction. Subsequently, a decision to access the spectrum
(e.g., frequency, bandwidth, modulation mode, transmission power, location, and time duration) is
made by optimizing the system performance given the desired objective (e.g., maximize throughput
of the unlicensed user) and constraints (e.g., maintain the interference caused to the licensed users
below the target threshold).
• Spectrum access: After the decision is made on spectrum access based on spectrum analysis, the
Page 8 of 66
http://mc.manuscriptcentral.com/wcm
Wireless Communications and Mobile Computing
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
For Peer Review
7
spectrum holes (also called spectrum opportunities) are accessed by the unlicensed users. Spectrum
access is performed based on a cognitive medium access control (MAC) protocol which intends to
avoid collision/harmful interference with/to licensed users and also with other unlicensed users. The
cognitive radio transmitter is also required to perform negotiation with the cognitive radio receiver
to synchronize the transmission so that the transmitted data can be received successfully. A cognitive
MAC protocol could be based on a fixed allocation MAC (e.g., frequency-division multiple access
[FDMA], time-division multiple access [TDMA], code-division multiple access [CDMA]) or a random
access MAC (e.g., ALOHA, carrier-sense multiple access with collision avoidance [CSMA/CA]).
The optimal spectrum access decision depends on the ambient environment and the cooperative or
competitive behavior of the unlicensed users. The spectrum access decision can be made in a non-
cooperative and distributed way based on a local optimization objective. Alternatively, a cooperative
spectrum access decision can be made either in a distributed or a centralized way based on a global
optimization objective. The spectrum access decisions are then communicated among cognitive radio
transmitters and receivers.
In a cognitive radio network, the secondary users may use either an interference control (or spectrum
underlay) or an interference avoidance (or spectrum overlay) approach to exploit the spectrum
opportunities. In the spectrum underlay approach, the secondary users transmit over the same spectrum
as the primary users as long as the interference caused to the primary users does not exceed a threshold
level. Therefore, such an approach requires a sophisticated power control scheme for secondary
transmitters. In the spectrum overlay approach, the secondary users need to have the knowledge about
spectrum holes so that the secondary users can exploit them ensuring that there is no interference
caused to the primary users. The interference avoidance approach is, therefore, more conservative
than the interference control approach, and no strict power control is required for this spectrum access
paradigm.
• Spectrum mobility: Spectrum mobility is a function related to the change of operating frequency band
of cognitive radio users. When a licensed user starts accessing a radio channel which is currently
being used by an unlicensed user, the unlicensed user can switch to the idle spectrum band. This
change in operating frequency band is referred to as spectrum handoff [9]. During spectrum handoff,
the protocol parameters at the different layers in the protocol stacks have to be adjusted to match
with the new operating frequency band. Spectrum handoff must ensure that the data transmission by
the unlicensed user can continue on the new spectrum band.
Page 9 of 66
http://mc.manuscriptcentral.com/wcm
Wireless Communications and Mobile Computing
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
For Peer Review
8
The above functionalities can be implemented based on the software-defined radio (SDR) which is a
reconfigurable wireless communication system. With SDR, the transmission parameters (e.g., operating
frequency band, modulation mode, and protocol) can be controlled dynamically. This adjustability function
is achieved by software-controlled signal processing algorithms. The main functions of SDR include multi-
band operation, multi-standard support, multi-service support, multi-channel support, and self-organization
(i.e., “plug-and-play” capability). The flexibility of SDR is useful for prototyping and evaluating cognitive
networking technology.
In a network of cognitive radios, the network infrastructure should support and facilitate the above
cognitive radio functionalities (i.e., spectrum sensing, analysis, access, and mobility). The network will
facilitate the communication among multiple cooperating cognitive radios, or the network will control
communication processes among competing cognitive radio users through proper radio resource allocation.
The policies specifying the limitations of radio access (e.g., available frequency bands in a certain location)
and the operating characteristics of the physical radio can be stored in the policy and configuration
databases which would be accessible to the cognitive radios through the network infrastructure.
III. ARCHITECTURES AND BUILDING BLOCKS OF COGNITIVE RADIO NETWORKS: A LAYERED
PERSPECTIVE
A dynamic spectrum access-based cognitive radio network may adopt either a cooperative or a non-
cooperative network architecture. In the non-cooperative CRN architecture, each cognitive node is respon-
sible for its own decision. Therefore, it has minimal communication requirements (hence less overhead);
however, the spectrum utilization may be low. On the other hand, in the cooperative but centralized CRN
architecture, a centralized server maintains a database of spectrum availability and access information
(based on information received from secondary users, e.g., through a dedicated control channel). Therefore,
spectrum management is simpler and coordinated in this case, and it enables efficient spectrum sharing.
For dynamic spectrum access, a cooperative but distributed CRN architecture relies on cooperative local
actions throughout the network to achieve a performance close to the global optimal performance. In this
architecture, cognitive radios periodically exchange information on their local environment, communication
requirements and performances among themselves. The cognitive radios use their local information as
well as information received from their peers to determine their communication parameters. The network
performance, however, may suffer from the hidden node problem and large control overhead. Note that
in both centralized and distributed strategies, the primary user may or may not cooperate. The protocol
Page 10 of 66
http://mc.manuscriptcentral.com/wcm
Wireless Communications and Mobile Computing
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
For Peer Review
9
architecture for a cognitive radio system is shown in Fig. 1. The communication protocols at the different
layers need to function such that the utilization of the radio spectrum is maximized while satisfying the
policy constraints.
Physical layer (PHY): In the physical layer (PHY), spectrum sensing is the primary task which includes
functionalities such as the following: detect spectrum holes/opportunities over wide frequency spectrum,
opportunity estimation, estimate interference at the primary receiver (e.g., in a CDMA environment).
Spectrum sensing involves obtaining the spectrum usage characteristics across multiple dimensions such
as time, space, frequency, and code (e.g., modulation, waveform, bandwidth, carrier frequency, etc). A
primary transmitter can be detected by using methods such as matched filter detection, energy detection,
and cyclostationary feature detection. The leakage power at the RF front-end of a primary receiver can
be measured/estimated to detect the activity of the corresponding primary receiver. In the physical layer,
the RF front-end is implemented based on software-defined radio (i.e., SDR transceiver). The RF front-
end requires high sampling rate, high resolution analog-to-digital converters (ADCs) with large dynamic
ranges, multiple analog front end circuitries, and high-speed signal processors to perform computationally
demanding signal processing tasks with small delay.
MAC layer: In the MAC layer, a decision is taken on whether to transmit or not considering that
the spectrum sensing/detection could be erroneous and how to exploit the spectrum holes (e.g., what
modulation and power level to use), and in case of transmission how to share the spectrum with other
cognitive radios. The MAC protocol functionalities include the following: obtain information on channel
occupancy (spectrum sensing1) and make decision on spectrum access; synchronize transmission param-
eters (e.g., channel and time slot) between transmitter and receiver; facilitate negotiation among primary
users and secondary users for spectrum allocation; facilitate communication among secondary users to
perform channel sensing and channel access; and facilitate spectrum trading functions (e.g., spectrum
bidding and spectrum pricing). The major challenges in designing the MAC protocols for CRNs include
optimal channel sensing for multi-channel access; primary users’ time-varying activity; hidden and exposed
terminal problems; synchronization between transmitter and receiver for which the unavailability of any
fixed common control channel and the channel availability changes depending on the spectrum access by
primary users; optimal channel allocation/scheduling, rate and power adaptation; and coexistence between
primary and secondary users.
1The spectrum sensing in the physical layer relates to signal detection and processing, while the spectrum sensing in the MAC layerinvolves the scheduling of a spectrum sensing activity.
Page 11 of 66
http://mc.manuscriptcentral.com/wcm
Wireless Communications and Mobile Computing
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
For Peer Review
10
Network layer: The primary tasks at the network layer are topology construction, addressing, and routing.
The topology construction involves spectrum detection, neighbor discovery, and topology management
(e.g., through spectrum mobility). The addressing can be fixed (e.g., extension of PHY/MAC address) or
dynamic (e.g., using dynamic host configuration [DHCP] protocol). In multi-hop CRNs, routing decisions
need to be made based on topology, MAC congestion, and link quality/reliability (a cross-layer solution).
Note that there is no fixed common control channel (e.g., for route discovery) and rerouting may need to
be performed frequently due to intermittent connectivity among cognitive nodes. The routing metric has
to consider parameters such as hop-count and link quality as well as spectrum management parameters
such as interference to primary users and availability of spectrum holes.
Transport layer: The transport layer protocol is responsible for flow and congestion control, the per-
formance of which is affected by MAC protocol performance and spectrum mobility. The throughput
performance of the traditional transmission control protocol (TCP) is a function of round-trip time
(RTT) and packet loss probability, which would depend on the spectrum management/sharing protocols,
bandwidth of spectrum holes, transmission power and interference levels. During rerouting, RTT and
packet loss rate change, and spectrum handoff latency (due to spectrum mobility) may increase the value
of RTT. Therefore, the performance of TCP in a CRN would be improved if the packet delay and loss
can be minimized during spectrum handoff (e.g., through efficient queue management).
The adaptive protocols in the MAC, network, transport, and application layers should be aware of the
variations in the cognitive radio environment. In particular, the adaptive protocols should consider traffic
activity of primary users, transmission requirements of secondary users, and variations in channel quality.
For this, the MAC-layer beacons can be sensed for information about sleep and wakeup schedules of the
cognitive radios and information about nodes’ connectivity. The routing protocol messages, for example,
can be exploited to detect mobility of the cognitive radio nodes. To link all the protocol modules in the
stack and enable cross-layer interaction, a cognitive radio control is used to establish the interfaces among
SDR transceiver, adaptive protocols, and wireless applications and services. This cognitive radio module
(Fig. 1) uses intelligent algorithms to process the measured signal from the physical layer, and to receive
information on transmission requirements from the applications to control the protocols parameters in the
different layers.
A programmable radio architecture and related protocol stack with cross-layer interactions as described
above along with the wideband radio front-ends will be the key enabling technologies for the cognitive
radio networks. In the following section, we survey the existing literature on PHY, MAC, and networking
Page 12 of 66
http://mc.manuscriptcentral.com/wcm
Wireless Communications and Mobile Computing
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
For Peer Review
11
Applications and services
Transmit/ receive
Cogn
itive
radi
o
SDR transceiver
MAC
Network
Transport
Adap
tive
prot
ocol
Fig. 1: Cognitive radio protocol stack.
layer issues in cognitive radio networks.
IV. CURRENT STATE-OF-THE-ART RESEARCH ON SPECTRUM SHARING IN COGNITIVE RADIO
NETWORKS
In this section, we provide a survey on the evolution of cognitive radio research covering different
aspects of spectrum sharing, i.e., spectrum sensing [41]- [61], measurements and statistical modeling of
spectrum usage [62]- [68]; physical layer (PHY) aspects such as waveform and modulation design [69]-
[92]; multiple access, resource allocation and power control, and spectrum mobility [93]- [125]; cognitive
learning, adaptation and self-configuration [126]- [140], and multi-hop transmission and routing [141]-
[148].
A. Spectrum Sensing, Interference Modeling, Measurements and Statistical Modeling of Spectrum Usage
Spectrum statistics are important information for secondary users to opportunistically access spectrum
of primary users. Therefore, there is a need to study and understand the characteristics and natures of
spectrum usage by the primary users. To gain such information, the following issues are considered in
the literature.
• Spectrum sensing: Spectrum sensing is a primitive action for secondary users to detect the status of
spectrum access by primary users. Without this information, the secondary users may not access idle
Page 13 of 66
http://mc.manuscriptcentral.com/wcm
Wireless Communications and Mobile Computing
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
For Peer Review
12
spectrum, reducing utilization, or may cause interference to the primary users occupying spectrum.
• Interference modeling: Secondary users may observe the interference on spectrum for two reasons.
Firstly, the secondary users must ensure that their transmission will not interfere and interrupt the
ongoing transmission of primary users. Secondly, given an interference condition, the secondary users
must access spectrum such that their transmission requirements are satisfied. Interference modeling
provides the secondary users with the capability to achieve these goals.
• Measurements and statistical modeling of spectrum usage: While spectrum sensing is a short-term
action to observe an instantaneous status of spectrum, spectrum measurement is performed in a long-
term basis (e.g., over a few months) to gain knowledge and statistical data of primary users. This
information is useful for the secondary users to determine their spectrum access strategy (e.g., access
in a particular time of the day to minimize interference to the primary users).
In the following, we provide reviews for the works related to these issues.
An overview of the regulatory requirements (e.g., sensing periodicity and detection sensitivity) and
the major challenges associated with spectrum sensing in cognitive radio networks are given in [41],
[42]. A comprehensive survey of the different spectrum sensing techniques can be found in [43], [44].
Traditional spectrum sensing techniques include the following: matched filter detection, energy detection,
cyclostationary detection, wavelet detection, and covariance detection. The performance metrics generally
used for spectrum sensing are: probability of correct detection and probability of false alarm. The main
challenges of spectrum sensing arise due to the uncertainty in a radio environment caused by channel
fading, interference, and noise, requirement of quick detection, and also due to the fact that the modulation
schemes, transmission powers, bit rate may be different for the different primary users. Also, a secondary
user may not be able to detect the location of the primary receiver. In this context, we refer to [46] for a
comprehensive overview of the different channel fading models. In [47], the authors evaluate the secondary
channel capacities under the average and peak interference power constraints at the primary receiver in
different fading environments such as Rayleigh fading, Nakagami fading, and log-normal shadowing. It
is shown that with the same interference power limit, channel capacity in fading environments exceeds
that of the AWGN channel.
In [48], the authors use random matrix theory to obtain the probability distributions of the test statistics
and find the closed-form expressions for the probability of detection and the probability of false alarm.
Also, the authors propose spectrum sensing methods which overcome the noise uncertainty problem, and
can be used without requiring the knowledge of signal, channel, and noise power. [49] studies the impact of
Page 14 of 66
http://mc.manuscriptcentral.com/wcm
Wireless Communications and Mobile Computing
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
For Peer Review
13
the uncertain information about the noise power level on the performance of the energy detector spectrum
sensing. The analysis of an energy detector with estimated noise power (ENP) is presented. This analysis
can be used to determine the threshold and establish conditions for existence of the SNR wall. In [51],
the authors provide a tutorial on the multi-taper method (MTM), which is a nonparametric method for
spectrum sensing. This method provides high spectral-resolution capability, estimates the average power
in each sub-band of the spectrum, and identifies the unknown directions of interfering signals. In [52], the
spectrum sensing duration is optimized to maximize the throughput of secondary users while achieving a
target detection probability in order to protect the primary users.
For improved performance, cooperative spectrum sensing can be used [53]- [60]. The cooperative
spectrum sensing process works as follows. Every cognitive radio performs its own local spectrum sensing
independently and then makes a binary decision on the presence of a primary user. Then, all of the CRs
transmit their decisions to a common node. The common node fuses the decisions and makes a final
decision to identify whether a primary user is present or not. The fusion techniques of cooperative spectrum
sensing can be either decision fusion or data fusion. Through cooperation, the reliability of primary user
detection can be improved. However, the performance of cooperative sensing can be limited when the
reporting channels are subject to fading and/or shadowing. In [53], the authors propose several robust
cooperative spectrum sensing techniques based on the concepts of cooperative diversity and multiuser
diversity.
In [56], the authors focus on the cooperative spectrum sensing system in which the individual cognitive
radio users make independent decisions about the presence of the primary signal and communicate their
decisions to a fusion center. Then, the fusion center makes the final decision about the occupancy of the
band. The observations at the nodes are assumed to be correlated, and therefore, a linear-quadratic (LQ)
fusion strategy is proposed as compared with counting rule which is linear combining. The proposed LQ
rule requires the second-order statistics of the local decisions when the primary users are present, which
can be obtained by estimating the pairwise statistics. In [58], the authors present optimal weightings for
linearly combining the energies measured at the cognitive radio users such that the probability of detection
is maximized with a constraint on the probability of false alarm. The amount of sensing and transmission
overhead in a cognitive radio network with large number of cooperating users and frequency channels
is large when every node scans all the channels and sends each channel’s status to the fusion center. To
reduce the communication overhead, [59] proposes matrix completion and joint sparsity algorithms for
spectrum sensing based on a small number of observations.
Page 15 of 66
http://mc.manuscriptcentral.com/wcm
Wireless Communications and Mobile Computing
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
For Peer Review
14
The work in [60] proposes a location-aware cooperative sensing algorithm which decides if a primary
user exists within a certain detection area or not. The primary network is modeled as a random geometric
network and the cognitive radio network is assumed to know the relative locations of its users. The optimal
weightings, that linearly combine the measured energy levels of the sensing nodes such that a modified
deflection coefficient is maximized, are obtained by Fisher linear discriminant analysis.
The work in [62] reports a comprehensive experimental study on interference analysis in the IEEE
802.22-based wireless regional area network (WRAN) system. The work in [42] is a significant contribution
towards modeling the distribution of aggregate interference at a primary receiver due to the secondary
transmitters in terms of system parameters of a spectrum sensing-based cognitive radio network. The
work in [63] is also an important contribution to statistical modeling of aggregate interference caused to
a primary user in a cognitive radio network. The theory of truncated stable distributions is used for the
modeling. The effect of power control on the cognitive network interference is also considered. [64] is
one of the very early studies on spectrum occupancy and interference in cognitive radio systems. [64]
discusses the opportunities, challenges, and communication limits in the cognitive radio technology. [65]
provides a framework for spectrum measurement and offline processing of measured data. The framework
has been evaluated by using on real-world spectrum measurement data. The authors in [66] propose a
statistical model for spectrum occupancy in time and frequency domain, and the key model parameters
are determined from actual measurements. The work in [67] presents a spectrum measurement campaign
in detail and the lessons learned in this campaign. Also, the paper presents a stochastic model based on
modified beta distribution for the duty cycle of occupancy in a sub-band. The authors in [68] address the
potential drawback of Poisson modeling for primary user activity and introduces a new model based on
the first-difference filter clustering and temporal correlation statistics.
There are a few design considerations among different spectrum sensing methods. Firstly, the cooperative
spectrum sensing can improve the detection performance, but at the cost of higher false alarm probability
and larger overhead (e.g., processing and communication). Spectrum sensing can be performed in a reactive
or proactive manner. A reactive spectrum sensing method will not incur an overhead if spectrum access is
not performed as in a proactive spectrum sensing. However, the reactive spectrum sensing will have longer
latency to collect radio environment status when a secondary user needs to access spectrum. Finally, the
tradeoff between rate and reliability must be optimized. Spending longer time in spectrum sensing will
lead to a more accurate sensing result. However, this is at the cost of lower rate due to shorter time for
transmission.
Page 16 of 66
http://mc.manuscriptcentral.com/wcm
Wireless Communications and Mobile Computing
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
For Peer Review
15
B. Waveform and Modulation Design for Cognitive Radios
To minimize interference to primary users, the design of waveform and modulation of signal from
secondary users can be optimized. For example, in an underlay spectrum access scheme, the secondary
user can transmit based on ultra-wideband (UWB) and adjust the pulse position and/or width to avoid
interference to the narrow band transmission of the primary users. Similarly, in an overlay spectrum access
scheme, the secondary user adapts the multicarrier modulation for an OFDM-based system to minimize
interference.
1) Waveform design and interference mitigation: In [69], a new pulse waveform design method is
proposed for the Ultra Wideband (UWB) telecommunication system which has to coexist with some
other narrowband systems. In [72], a novel low-complexity adaptive UWB pulse shaping algorithm is
presented for producing the expected spectral notches right in the frequency band occupied by the nearby
wireless devices. Adaptive pulse waveform optimization based on linear combination of different derivative
functions of Gaussian pulse is investigated to produce expected notches in achieving narrowband inter-
ference avoidance as well as matching with the FCC spectral mask. In [75], non-contiguous multicarrier
modulation that can be used to facilitate spectrally opportunistic cognitive radio, is considered. [76] jointly
designs the transmitted waveform and the FIR filters at the relay nodes for cognitive relay networks. [77]
considers the problem of waveform optimization for multiple input single output cognitive radio links with
the spectral mask constraint at the transmitter and the interference cancellation at the receiver. In [86],
a novel method of adding extended active interference cancellation signals to suppress sidelobes and to
shape the spectrum of CR-OFDM signals with a cyclic prefix (EAIC-CP) is presented. The key idea of
the proposed EAIC-CP method is to use the cancellation signals consisting of EAIC tones that are spaced
closer than the OFDM subcarrier interval.
In [87], a statistical model of interference aggregation in spectrum sensing-based cognitive wireless
networks is presented by explicitly taking into account the random variations in the number, location
and transmitted power of the cognitive radios as well as the propagation characteristics. [90] proposes
an enhanced opportunistic interference mitigation scheme utilizing both successfully and unsuccessfully
decoded primary packets to improve the data rate of secondary user’s transmission. Also, an analytical
model is proposed to investigate the overall spectral efficiency due to the changes of spectrum usage
affected by the coexisting secondary user’s transmission.
2) Modulation design and classification: [91] provides an excellent tutorial on two multicarrier mod-
ulation schemes, namely, non-contiguous orthogonal frequency-division multiplexing (NC-OFDM) and
Page 17 of 66
http://mc.manuscriptcentral.com/wcm
Wireless Communications and Mobile Computing
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
For Peer Review
16
non-contiguous non-orthogonal frequency-division multiplexing (NC-NOFDM) for opportunistic wireless
access. [78] also reviews different modulation schemes (e.g., DFT-spread OFDM, constant envelope OFDM
and filter bank multicarrier) for cognitive radios with their advantages and disadvantages. Different mod-
ulation schemes are compared in terms of spectral characteristics, signal metrics, and system performance
with the presence of non-linear distortions. In [84], a quantitative comparison of NC-OFDM and non-
contiguous multicarrier CDMA transmission techniques is presented in the context of a dynamic spectrum
access network. In [85], the authors provide a review of the state-of-the-art methods specifically designed
to protect the transmissions of the primary users from possible interference caused by nearby secondary
transmitters using NC-OFDM. These methods are designed to suppress out-of-band (OOB) emissions
resulting from the use of NC-OFDM transmission.
In [79], the performance of adaptive modulation for cognitive radio with opportunistic access is analyzed
by considering the effects of spectrum sensing and primary user traffic for Nakagami-m fading channels.
Both the adaptive continuous-rate scheme and the adaptive discrete-rate scheme are considered. [83]
introduces modulation strategies to realize the coexistence between the CR-based rental system and the
licensed system in such a way that the secondary users are invisible to the primary users.
In [89], the techniques for detecting interfering signals, estimating the distance between victim and
interfering devices, and shaping the spectrum to mitigate interference into the victim receiver are con-
sidered. [80] designs a classifier for linear digital modulation based on the clustering (using the K-
Means algorithm) of the received signal constellation diagram. [82] presents a novel digital modulation
classification system for cognitive radios using only temporal waveform features. A hierarchical approach
is used to first make separations into intermediate subclasses, where some of the subclasses can consist
of more than one modulation type. Then the second classifier is used to discriminate between higher
order modulation schemes using additional features. In [92], the authors propose a method for modulation
classification based on the Support Vector Machines (SVMs) which is robust against heavy non-Gaussian
interference. In the experiment, four classes of signal modulations, i.e., binary phase-shift keying (BPSK),
quadrature phase-shift keying (QPSK), frequency-shift keying (FSK), and minimum shift keying (MSK),
are considered.
C. Analysis of Spectrum Sharing Systems
Capacity analysis is still an important issue to quantify the upper bound performance of the cognitive
radio system. The capacity of a cogntiitve radio system is constrained not only by the transmit power
Page 18 of 66
http://mc.manuscriptcentral.com/wcm
Wireless Communications and Mobile Computing
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
For Peer Review
17
limit, but also by the interference to primary users. This analysis is useful to determines the achievable
transmission rates of secondary users. The analysis can consider the effects of different channel fading.
[47] evaluates the secondary channel capacities under the average and peak interference power con-
straints at the primary receiver in different fading environments. It is shown that with the same interference
power limit, channel capacity in fading environments exceeds that of the AWGN channel. The impacts
of correlated fading and multiple primary receivers on the channel capacity are also studied. Asymptotic
capacities under different fading distributions are derived. The work in [93] analyzes the ergodic capacity of
secondary user system for different spectrum sharing strategies (underlay, overlay, and hybrid strategies)
with one primary user and one secondary user. The analysis reveals that the maximum capacity in a
spectrum underlay system can be achieved with reduced signaling complexity where the channel state
information between the secondary transmitter and the primary receiver may not be required for power
allocation.
In [94], the authors discover a new way of spectrum sharing by exploiting the primary Automatic
Repeat and reQuest (ARQ) retransmissions for secondary user communications. The paper also avoids
somewhat impractical assumptions such as complete CSI at the secondary transmitter. [110] presents a
learning-based spectrum sharing protocol for spectrum-sharing wireless networks such that the weighted
time-fairness can be achieved among the different networks sharing the spectrum.
D. Multiple access, Resource Allocation, and Power control
1) Multiple access in spectrum underlay: In a spectrum underlay scenario, the problem of optimal
spectrum sharing among secondary users can be formulated as an optimization problem with a suit-
able objective function and a set of constraints which capture user fairness, quality-of-service (QoS) of
secondary users, and interference constraints for primary users as follows:
max U(R1, R2, . . . , Rn)
subject to
Ri ≥ Bi, i = 1, 2, . . . , n
µj =n∑
i=1
hj,iPi ≤ Ij, j = 1, 2, . . . ,m
where U is the objective function, Ri is the achievable rate for secondary user i, Bi is the minimum
rate requirement for user i, µj is the total interference created by secondary users at primary receiver j,
Page 19 of 66
http://mc.manuscriptcentral.com/wcm
Wireless Communications and Mobile Computing
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
For Peer Review
18
hij denotes the channel gain from the transmitter of secondary user i to primary receiver j, Pi denotes
transmission power of secondary transmitter i, and Ij denotes the maximum tolerable interference level
at primary receiver j.
Note that the optimization problem above may not be feasible when its constraints are too stringent
and/or the network load is too high. If this is the case, an admission control mechanism needs to be
invoked to limit the number of admitted secondary users. Then, the power allocation for the set of admitted
secondary users can be performed. Using this framework, in [96], a solution approach is proposed for
the joint admission control and power allocation problem for secondary users to achieve fairness among
them assuming CDMA technology at the physical layer. A related spectrum sharing problem, which
maximizes throughput of secondary networks, is solved in [97]. To obtain the power allocation solutions,
the instantaneous channel gains among secondary users and interference from secondary transmitters
to primary receivers need to be estimated. In a practical scenario where only estimates of the average
channel gains are available, the power allocations for the secondary transmitters need to be done in a
conservative way to satisfy the target interference constraint violation probability for primary receivers [98].
The problem of sum-rate maximization for secondary transmitters under joint beamforming and power
allocation in a cognitive radio network with multiple secondary transmitters and primary receivers (each
with one antenna) is investigated in [100].
The performance of a cognitive radio network could be enhanced by using multiple antennas at the
secondary transmitters and/or secondary receivers. However, in this case, power allocations among the
transmit antennas under total transmission power constraints and interference-power constraints at the
receiver antennas need to be jointly considered. Also, the tradeoff between throughput and fairness in a
multiuser MIMO cognitive radio network needs to be modeled and analyzed. The work in [99] exploits
multiple antennas at the secondary transmitter to maximize the cognitive radio’s transmission rate under
the transmission power constraint and a set of interference power constraints. The authors demonstrate the
effectiveness of using convex optimization for solving resource allocation problems for cognitive spectrum
sharing.
A survey on dynamic resource allocation schemes for cognitive radio systems with the interference
temperature based spectrum sharing model can be found in [101]. To maximize the secondary network
throughput, the transmission power, bit-rate, bandwidth, and antenna beam can be dynamically allocated
based upon the available channel state information (CSI) of the primary and secondary networks. Many
new and challenging problems regarding the design of CR systems are formulated, and some of the
Page 20 of 66
http://mc.manuscriptcentral.com/wcm
Wireless Communications and Mobile Computing
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
For Peer Review
19
corresponding solutions are shown to be obtainable by restructuring some classic results known for
traditional (non-CR) wireless networks.
2) Multiple access in spectrum overlay: In a spectrum overlay scenario, secondary users need to sense
a frequency band of interest and transmit only if primary users are not detected on the chosen band.
Therefore, no strict power control is required for this spectrum access paradigm. Either scheduling or
random access-based multiple access techniques can be used [102], [103]. In the former case, a scheduler
(e.g., in a base station or an access point) makes the scheduling decisions based on the spectrum sensing
results and/or the statistical information regarding channel availability. Generally, either a Markovian
or an independence assumption is made for channel availability. In the latter case, the secondary users
need to discover the spectrum opportunities to access the channel(s) through a contention resolution and
reservation process. The hardware constraints (i.e., partial spectrum sensing and spectrum aggregation
limit) of secondary users need to be considered for a practical channel access protocol.
The work in [104] studies the problem of optimal discovery of spectrum opportunities with MAC-
layer sensing in cognitive radio networks. The objectives are to maximize the discovery of spectrum
opportunities by adapting sensing periods and to minimize the delay in finding an available channel. The
analysis defines sensing overhead and unexplored opportunity, and optimizes these factors. The sensing
overhead is the average fraction of time during which a channel’s discovered opportunities are interrupted
and not utilized due to sensing of one of the channels. The unexplored opportunity is the average fraction of
time during which channel’s opportunities are not discovered in case that the channel is being periodically
sensed with its sensing period. The optimal channel sequencing can be obtained from the probability that
the channel would be idle at a certain time based on the previous samples. The proposed approach requires
the estimation of the underlying channel-usage patterns with an ON/OFF alternating renewal channel. The
distribution parameters of ON/OFF periods are estimated by Maximum Likelihood (ML) estimators and
their confidence intervals.
With a Markovian assumption on channel availability, the work in [105] proposes decentralized cognitive
medium access method in a single-user scenario which allows a secondary user to independently search
for spectrum opportunities without a central coordinator or a dedicated communication channel. An
analytical framework is developed based on the theory of Partially Observable Markov Decision Process
(POMDP). However, the optimal policy for a general POMDP can be computationally prohibitive due to
the exponential growth of the dimension of the sufficient statistic. The authors propose a greedy approach
that maximizes per slot throughput. This decision-theoretic approach integrates the design of spectrum
Page 21 of 66
http://mc.manuscriptcentral.com/wcm
Wireless Communications and Mobile Computing
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
For Peer Review
20
access with spectrum sensing at the physical layer and traffic statistics determined by the application layer
of the primary network. An opportunistic user makes optimal decisions for sensing and access based on
the belief vector that summarizes the knowledge of the network state based on all past decisions and
observations.
With an independence assumption on channel availability, [106] proposes a hardware-constrained mul-
tichannel cognitive MAC protocol considering the tradeoff between exploring more idle channels and
encountering more sensing overhead. The channel access problem is formulated as a well-defined optimal
stopping problem. The objective is to choose a time to stop spectrum sensing (probing) such that the
expected reward is maximized. The decision is either to stop and receive the known reward or to continue
and observe the channel status (i.e., the 0-1 (occupied/idle) state of the channel probed) for further decision.
This problem can be solved by the backward induction method. The similar case of correlated traffic from
primary users is also considered in [107].
3) Cross-layer multiple access: In [108], the authors characterize quantitatively the interaction between
the physical and MAC layers. In particular, they demonstrate how the sensing errors at the PHY layer affect
the MAC design and how incorporating the MAC layer information into physical layer leads to a cognitive
spectrum sensor whose performance improves over time by learning from accumulating observations.
The cross-layer channel access problem is formulated as a constrained POMDP. It is shown that there
exists a separation principle in the optimal joint design of channel access that circumvents the curse of
dimensionality in general POMDPs and leads to closed-form optimal solutions.
[111] proposes a cross-layer based opportunistic channel access protocol that integrates spectrum
sensing at the physical layer and packet scheduling at the MAC layer. The authors assume that each
secondary user is equipped with two transceivers. One transceiver is tuned to the dedicated control channel
and another transceiver is used to periodically sense and dynamically use the identified unused channels.
Two channel spectrum-sensing policies, namely, the random sensing policy and the negotiation-based
sensing policy, are considered. Under the random sensing policy, each secondary user randomly selects
one of the channels for sensing with a uniform distribution. Under the negotiation-based sensing policy,
each secondary user attempts to select a channel for sensing by overhearing the control packets in the
control channel. The authors develop an M/GY /1-based queuing model to analyze the performance of
the proposed multi-channel MAC protocols under the two types of channel-sensing policies for both the
saturation and non-saturation conditions. The analyses reveal the tradeoff between throughput and delay,
which would be useful to support the different QoS requirements in cognitive radio networks.
Page 22 of 66
http://mc.manuscriptcentral.com/wcm
Wireless Communications and Mobile Computing
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
For Peer Review
21
In multichannel cognitive radio networks, one important issue is how a secondary user chooses the
sequence of channel sensing such that the best idle channel can be identified within the shortest time
period, maximizing throughput of the secondary user. [112] considers the optimal channel sensing order
by formulating and solving the dynamic programming model. This optimal sensing order is part of a
cognitive medium access control (MAC) which can be implemented given a spectrum sensing technique
in a physical layer. In the simplest case, [112] first shows that the intuitive sensing order (i.e., descending
order of the channel availability probabilities) is optimal. Then, the model is extended for the system with
adaptive modulation transmission, when the probability of an idle channel is known. Finally, the most
complex scenario, where the probability of an idle channel is unknown and there is a sensing error, is
also considered.
Cross layer design and optimization can be applied for multiple access of cognitive radio nodes. In
this case, the MAC protocol can utilize information from protocols in other layers (e.g., physical and
network layers) to optimize its performance. However, the cross layer approach may suffer from losing
independency of a protocol in each layer. That is, an operation of a protocol in one layer depends on a state
of protocols in other layers. Nevertheless, a few works study the benefits of cross layer design in cognitive
radio networks. [113] integrates spectrum sensing in the physical layer with the MAC protocol. Specifically,
the collaborative spectrum sensing policies are introduced, i.e., random sensing and negotiation-based.
While the former is a simple policy where secondary users can decide on spectrum sensing and access
randomly and independently, the latter allows the coordination so that the secondary users choose different
channels to sense and access. [114] studies the use of location information of a primary user to optimize
the PHY-MAC protocol of secondary users. The location information is obtained through observing RSSI
and direction of signal from an antenna array. The collaborative sensing algorithm is proposed to utilize
this location information and optimize the spectrum access given the interference constraint to the primary
user. The similar problem is also studied in [115] with the reporting overhead taken into account.
4) Power control: In [116], the authors discuss the coexistence of the primary and cognitive users in the
same spectrum by using power allocation strategies at the cognitive radios such that the interference power
experienced by the primary receiver is limited. The work considers various combination of peak/average
transmission and interference power constraints, and studies the power allocation strategies to achieve
the ergodic, delay-limited, and outage capacities of the cognitive radio system. Different fading channel
models such as Rayleigh, Nakagami, and log-normal shadowing are considered.
The orthogonal frequency division multiplexing (OFDM)-based transmission technology is particularly
Page 23 of 66
http://mc.manuscriptcentral.com/wcm
Wireless Communications and Mobile Computing
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
For Peer Review
22
promising for cognitive radio networks using spectrum overlay. However, OFDM transmission causes
mutual interference between the primary and secondary users due to the non-orthogonality of the trans-
mitted signals. The amount of interference depends on the power allocated in the subcarriers as well
as the spectral distance between subcarriers used by the primary and secondary users. In [117], the
authors investigate the optimal power loading problem to maximize the transmission data rate while
maintaining the interference caused to the primary users within a given limit. The work is extended
in [118] considering different statistical interference constraints imposed by different primary receivers
where the probability that the interference introduced by secondary user’s transmission in the ith subcarrier
to the lth primary receiver (∑N
i=1 I(l)i (dil, Pi)) is less than the interference limit (I(l)th ) must be greater than
a (i.e., Prob(∑N
i=1 I(l)i (dil, Pi) ≤ I
(l)th
)≥ a).
[120] considers the problem of energy-efficient power allocation for maximizing the expected trans-
mission rate for OFDM-based cognitive radio systems. The reliability of the available sub-bands (which
depends on sensing error and primary user’s activity), sub-band power constraints, and total allowed
interference limit to the adjacent primary user bands are taken into account. The authors consider an
energy-aware capacity expression through the factor called subcarrier availability. A convex optimization
problem is formulated which incorporates a linear average rate loss function in the optimization objective
to include the effect of subcarrier availability. The objective function is to maximize the capacity minus
cost due to transmit power.
Due to the complex structure of the optimal solution, the authors propose three suboptimal schemes,
namely, the step-ladder, nulling, and scaling schemes, to redistribute the power within the sub-bands
which share boundaries with the primary users’ bands. The step-ladder scheme assigns lower power to
subcarriers closer to primary users’ bands with a step size which could be either constant or proportional
to the interference caused to a primary user. For the nulling mechanism, zero power is allocated to
subcarriers adjacent to the primary user band (one-nulling) or zero power is allocated to two subcarriers
closest to primary user band on each side (two-nulling). Finally, the scaling scheme allocates the power
that satisfies the interference criterion strictly by scaling down the power to the function on the total
interference threshold prescribed by primary user bands. The worst-case complexity of the algorithm is
O(MN), where M is the number of groups of the available sub-bands and N is the available subcarriers
to the secondary user for opportunistic access using OFDM.
[121] extends the traditional scheduling and power control problems by considering their robustness for
cognitive spatial-reuse time division multiple access (STDMA) network. The objective is to minimize the
Page 24 of 66
http://mc.manuscriptcentral.com/wcm
Wireless Communications and Mobile Computing
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
For Peer Review
23
transmission length (in terms of time slots) of the secondary links under their minimum QoS requirements
without violating the maximum tolerable interference limit for the primary receivers. Due to the stochastic
nature of wireless link gains, the power allocation based on average (or estimated) link gains can be
improper. To overcome the improper power allocation, the channel gain uncertainty is modeled by ellipsoid.
Since an optimal solution cannot be obtained due to the NP-completeness of the problem, the authors
propose a distributed two-stage algorithm based on the distributed column generation method to obtain
the near-optimal solution for the robust transmission schedules in an ad-hoc cognitive radio network.
E. Distributed Learning, Adaptation, Self-configuration, and Coexistence
Learning, adaptation, and self-configuration are the fundamental functions of cognitive radio. Secondary
users should be able to observe spectrum state and learn the status of a primary user, then optimize and
adapt the transmission so that the secondary users can achieve their performance target while minimizing
interference to the primary user. Self-configuration is also an important functionality for the secondary
users to determine a set of optimal system parameters. The self-configuration capability generally relates
to a long-term setting (e.g., frequency band and transmission range), while the dynamic spectrum access is
concerned about a short-term action (e.g., transmit power and time slot to access). The main aim of which
is to ensure that a secondary user can coexist with primary users and other secondary users amicably.
The work in [126] describes the motivations, architecture, functionality, design and implementation of
cognitive networks applicable to both wired and wireless networks. A cognitive process to learn from past
decisions and use this learning to influence future behavior is the foundation for such networks.
In [127], the authors propose a game theoretic framework to analyze the behavior of cognitive radios
for distributed adaptive channel allocation. The objective is to choose the frequency channels which
maximize the reward function. Two different reward functions are formulated to capture the utility of
selfish users (U1) and cooperative users (U2), respectively. U1 values a channel based on the level of
interference perceived on that particular channel while U2 accounts for the interference seen by a user on
a particular channel, as well as the interference this particular choice will create to neighbouring nodes.
The authors demonstrate that the channel allocation problem can be formulated as a potential game and
thus it converges to a deterministic channel allocation Nash equilibrium point. Although the game with
the utility function given by U2 fits the framework of an exact potential game, the function U1 lacks
the necessary symmetry properties that will ensure the existence of a potential function. To analyze the
behavior of the selfish user game, the authors resort to the implementation of adaptation protocols using
Page 25 of 66
http://mc.manuscriptcentral.com/wcm
Wireless Communications and Mobile Computing
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
For Peer Review
24
no-external-regret learning algorithm with exponential updates as described below.
By the proposed learning adaptation process, the CR users learn how to choose the channels that max-
imize their rewards through repeatedly playing of the game. The no-regret learning provides performance
similar to that of the potential game when cooperation is enforced, but with a higher variability across users.
However, the learning formulation is particularly useful to accommodate selfish users. Non-cooperative
learning games have the advantage of very low overhead for information exchange in the network while the
cooperative learning game improves the overall network performance at the expense of increased overhead
required for information exchange. Note that the case of non-convex game is considered in [128].
In [129], the authors provide a comprehensive survey on the use of artificial intelligence for cogni-
tive radio networks. It reviews different implementations of cognitive radio which are designed based
on intelligent algorithms including artificial neural networks (ANNs), metaheuristic algorithms, hidden
Markov models (HMMs), rule-based systems, ontology-based systems (OBSs), and case-based systems
(CBSs). [129] also discusses the issues related to the selection of intelligent algorithms (e.g., responsive-
ness, complexity, security, robustness, and stability) for cognitive radio networks.
The work in [130] introduces an innovative method of designing cognitive radio by adopting the human
behavior model. The concepts of homo egaulis society, homo parochius society, and homo reciprocans
society in social science are discussed in the cognitive radio context with illustration. The work in [131]
uses the machine learning techniques in cognitive radio, in contrast to the traditional methods which rely
on the policy-based and hard-coded approaches and presents a concrete model of a generic cognitive radio
with learning engine.
The work in [132] considers the problem of aggregated interference management due to multiple
cognitive radios using ideas from artificial intelligence. The IEEE 802.22 standard is modeled as the
multiagent system where agents are the secondary base-stations supporting and controlling the data trans-
mission of secondary users. A real-time multiagent reinforcement learning algorithm (i.e., decentralized
Q-learning algorithm) is introduced to manage the aggregated interference. Both complete and partial
information cases are evaluated in which the optimal and suboptimal solutions are obtained, respectively.
The computational and memory requirements for implementations of these reinforcement algorithms are
evaluated.
The work in [133] adopts the concept of Bayesian nonparametric learning algorithm to optimize the
repeated spectrum auction in a cognitive radio network. The repeated spectrum auction is used which
considers the monitoring and entry costs of the secondary users to bid for the radio resource from the
Page 26 of 66
http://mc.manuscriptcentral.com/wcm
Wireless Communications and Mobile Computing
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
For Peer Review
25
primary users. Since the knowledge of the secondary users about other users is limited due to the distributed
environment, a secondary user learns from experience using Bayesian nonparametric belief update scheme
and adapts the bidding strategies accordingly. Based on the belief, the secondary user decides to join the
auction or not. In addition, the bidding strategy is proved to be optimal.
In [135], the authors consider the multi-user distributed dynamic spectrum access problem in a spectrum
overlay scenario based on game-theoretic learning. In particular, a carrier-sense multiple access (CSMA)
protocol is considered and a game theoretic correlated equilibrium is achieved. In [136], the authors
consider cognitive radio networks for delay-critical applications using game theory and learning algorithm.
Previously, the cognitive radio is believed to be useful only for non-real-time applications (e.g., best-effort
applications). However, by using advanced techniques in optimization and artificial intelligence, this paper
proves the feasibility of deploying real-time applications (e.g., multimedia) in cognitive radio networks.
This paper presents centralized and decentralized spectrum market models based on the stochastic game
framework.
The work in [137] proposes the use of neural network from the field of artificial intelligence to implement
dynamic spectrum selection in cognitive radio networks. A neural network-based cognitive engine is
developed to learn from the environmental measurement on the different channels. Then, the algorithm
chooses the channel with the highest chance to yield the best performance based on the result from the
cognitive engine. The experimental measurement based on the testbed implementation is conducted to
evaluate the performance of the proposed cognitive radio design.
The work in [138] also adopts a neural network to predict the availability of the spectrum for cognitive
radio users. With a multilayer perceptron (MLP) neural network, where the training data is assumed to be
available, the accuracy of the spectrum prediction by the secondary user can be improved significantly.
Therefore, a secondary user can efficiently access the available spectrum without collision with the
transmission by the primary users. The authors in [139] investigate on distributed learning of channel
availability statistics and channel access in cognitive radio networks and present distributed channel access
policies based on the results on classical multi-armed bandit problem. For these policies, the bounds on the
regrets are obtained. The work in [140] presents a novel learning-based approach for dynamic spectrum
access by secondary users in a spectrum overlay scenario where the probability of collision with the
primary users needs to be bounded. This approach significantly outperforms the traditional listen-before-
talk approach.
Page 27 of 66
http://mc.manuscriptcentral.com/wcm
Wireless Communications and Mobile Computing
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
For Peer Review
26
F. Multi-hop Transmission, Routing, Spectrum Mobility, and Cross-layer Optimization
In multi-hop cognitive radio networks [141], cognitive radio users communicate with each other in
a local area either in a centralized or an ad-hoc manner. Hence, routing selection plays an important
role in multi-hop cognitive networks. A local coordination based routing scheme for multi-hop cognitive
networks is proposed in [142]. In [143], joint routing, subband division and scheduling scheme for multi-
hop CRNs is considered. In [145], a cross-layer optimal scheduling algorithm is proposed for cooperative
multi-hop cognitive radio networks. In this network, secondary users help primary users to transmit in a
multi-hop fashion and the secondary users receive a spectrum opportunity for their transmission in return.
The analysis is developed to achieve optimal throughput for the primary users, where the upper bound is
derived.
To exploit the time-varying channel opportunities and provide QoS requirement to the users and save
transmission power, the authors in [148] consider dynamic rate allocation, routing and spectrum sharing
scheme for multi-hop cognitive networks.
In [143], the authors study the problem of multi-hop networking with cognitive radio nodes. For the
cognitive radio network, each node has a pool of frequency bands (typically of unequal size) which can be
used for data transmission. The difference in the bandwidth among the available frequency bands provokes
the need to divide these bands into sub-bands for optimal spectrum sharing. The authors characterize the
behaviors and constraints for a multi-hop cognitive radio network from multiple layers, including the
modeling of spectrum sharing and sub-band division, scheduling and interference constraints, and flow
routing. The objective is to minimize the required network-wide radio spectrum resource for a set of user
sessions. The problem can be formulated as a mixed-integer non-linear programming (MINLP). However,
this MINLP problem is NP-hard. Therefore, the authors develop a lower bound to estimate the objective
function by relaxing the integer variables and using a linearization technique. Simulation results show
that the solutions obtained by the proposed algorithm are very close to the lower bounds obtained via the
proposed relaxation. Therefore, the solution produced by the algorithm is nearly optimal.
The work in [122] studies the data link layer QoS performance measures for cognitive radio users in
an infrastructure-based dynamic spectrum access environment. The authors develop a queuing analytic
framework with a discrete-time Markov chain (DTMC) and derive important QoS measures. The model
considers primary users’ activity as a two state Markov chain, cognitive radio users’ channel activity as
a finite state Markov chain, bursty traffic arrival pattern at the cognitive radio user ends, and correlated
Page 28 of 66
http://mc.manuscriptcentral.com/wcm
Wireless Communications and Mobile Computing
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
For Peer Review
27
channel fading. The authors provide a step-by-step procedure to derive the delay distribution, average
throughput, and packet loss rate for the cognitive radio users. The delay distribution can be derived by
considering an absorbing Markov chain. The proposed framework has important applications including
cross-layer analysis and design for improved QoS experience, and model-based admission control in
cognitive radio networks.
There have been only a few work on spectrum handoff and its effect on the performance of cognitive
radio networks. In [123], the authors outline the research issues and challenges related to spectrum
mobility in the context of spectrum management process for cognitive radio networks. The work in [124]
comprehensively models the effects of spectrum handoff and spectrum management methods on the call-
level QoS performance of secondary users. The work in [125] demonstrates spectrum handoff and spectrum
mobility features of cognitive radio (based on non-contiguous multi-carrier transmission) using software-
defined radio.
Robustness and security are important issues in the development of cognitive radio networks. In the
next section, we review the existing literature on security and robustness in cognitive radio networks.
V. ROBUSTNESS AND SECURITY OF COGNITIVE RADIO NETWORKS
A cognitive radio network could be vulnerable to different types of attacks, which could degrade
the performance considerably. [149] provides a comprehensive study and analysis on the security issues
in a cognitive radio network. The cognitive radio devices could wrongly learn from environment and
be taught by malicious users. Various attacks (e.g., sensory manipulation attacks against policy radios,
belief manipulation attacks against learning radios, and self-propagating behavior leading to cognitive
radio viruses) can use this approach. To address the problem, different methods are discussed. Firstly,
the robust sensory input can be performed by allowing the secondary users to run a cross-correlation
algorithm to determine the difference between noise and malicious signal. Secondly, all possible states of
the environment must be enumerated to obtain the complete knowledge so that the secondary user will
be aware of an attack. Thirdly, multiple secondary users can cooperate to form a group and exchange
information if there is an attack. Each of these methods may have some drawbacks. The cross-correlation
algorithm could be too complicated to implement in a typical secondary node. State enumeration requires
too much memory and too long execution time of a secondary user. Finally, forming a group and
exchanging information could incur an unacceptable overhead in a network. Alternatively, some defense
mechanisms for cognitive radios against the attack (e.g., using trust network) are also mentioned.
Page 29 of 66
http://mc.manuscriptcentral.com/wcm
Wireless Communications and Mobile Computing
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
For Peer Review
28
In a cognitive radio system, it is important to distinguish between the signals from primary and secondary
users. Especially, in a hostile environment, the attacker may use a modified air interface to imitate the signal
of primary users, making secondary users unable to access the spectrum. [150] systematically shows that
the aforementioned attack can result in severe interference and significantly reduce spectrum utilization.
To solve the problem, [150] proposes a transmitter verification scheme, namely, the LocDef (localization-
based defense) scheme. This LocDef scheme is able to identify whether the signal is from a primary
user or not by using an estimated location of the transmitter (i.e., based on non-interactive localization
technique) and characteristics of the signal itself (i.e., received signal strength [RSS]). [150] considers the
log-loss signal propagation model, in which the expected RSS is given by µ = P + β0 + β1 ln d, where d
is the distance between actual primary transmitter and secondary receiver, P is the transmission power,
and β0 and β1 are constants for a particular environment. The value of RSS is then smoothed to extract
the important pattern, and the RSS peak can be used to identify the primary transmitter.
Adversaries
Emulated incumbent signal
Local sensing reports
Attacker
Falsified sensing result
Incumbent signal transmitter
(primary user) Data collector
Actual incumbent
signal
Final spectrum sensing result
Spectrum sensors
Fig. 2: Spectrum emulation attack.
In [151], the authors consider incumbent emulation and spectrum sensing data falsification attacks
(Fig. 2) which can degrade the performance of distributed spectrum sensing in cognitive radio networks.
In particular, an attacker (e.g., malicious secondary user) tries to gain priority over other secondary
users by transmitting signal emulating that of a primary user (i.e., incumbent emulation) or reporting false
spectrum sensing data (i.e., spectrum sensing data falsification). Defense methods against these attacks are
introduced. For the incumbent emulation attack, defense methods based on distance ratio test and distance
difference test are considered. The distance ratio test uses the RSS obtained from the pair of location
verifiers (e.g., dedicated spectrum sensor devices or secondary user devices) to verify the location of the
transmitter. However, the distance ratio test suffers from channel fading and shadowing, and may not yield
Page 30 of 66
http://mc.manuscriptcentral.com/wcm
Wireless Communications and Mobile Computing
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
For Peer Review
29
the accurate distance to the transmitter. The distance difference test uses the relative phase difference of
the signal obtained at a pair of location verifiers to verify the transmitter. The distance difference test needs
synchronization among location verifiers to obtain the accurate result. For the spectrum sensing falsification
attack, the attacker tries to manipulate the spectrum sensing at the spectrum sensors or data collector (i.e.,
fusion center). Different methods related to the data fusion techniques are considered against the spectrum
sensing falsification attack, i.e., decision fusion, Bayesian detection [152], and Neyman-Pearson test. In
the decision fusion method, the sensing results from spectrum sensors are compared with the rule (e.g.,
threshold-based). In the Bayesian detection, the knowledge of a priori conditional probabilities of the
spectrum sensing is used to detect the attack. In the Neyman-Pearson test, the maximum acceptable
probability of false alarm or the maximum acceptable probability of misdetection is used to determine
the presence of an attack.
The work in [153] and [154] develop a defense strategy against the primary user emulation attack. The
passive anti-primary user emulation attack approach here is similar to the random frequency hopping used
for anti-jamming in traditional wireless networks. In this approach, the defender randomly switches the
channel to sense and access so that the defender can escape from the attack on a particular channel by the
attacker. In [153] (unlike that in [154]), the channel statistics (e.g., availability probability) are assumed to
be known. In [153], the attack and the defense mechanisms are modeled as a zero-sum game. The actions
of the attacker and the defender are the channel to jam and to perform channel sensing, respectively.
The payoff of the attacker is a constant reward if the attacker successfully attacks the defender, and
zero otherwise. Similarly, the payoff of the defender is the same constant reward if the defender can
successfully avoid the attack by the attacker, and zero otherwise. The Nash equilibrium is analyzed for
this game. The metric named anti-jamming efficiency (AJE) is introduced to measure the outcome of the
game, which is defined as follows:
AJE =Expected reward at the Nash equilibrium
Expected reward without jamming. (1)
The exact expression for anti-jamming efficiency is derived. The analysis in [153] also considers the
multiple stage case. The problem is simplified due to the partial observation of channel sensing and
imperfect monitoring. A POMDP is formulated and solved for the secondary user defender to optimally
sense and access the channel. The work in [154] extends the analysis of a primary user emulation attack
for the unknown channel statistics case. The secondary user as the defender must address the uncertainty
Page 31 of 66
http://mc.manuscriptcentral.com/wcm
Wireless Communications and Mobile Computing
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
For Peer Review
30
in the channel statistics and time-varying policy of the attacker. The defender therefore has to learn and
adapt the channel selection policy accordingly.
In [155], the authors consider the problem of malicious secondary users deliberately sending false
reports in collaborative spectrum sensing. This is the spectrum sensing falsification attack. In a cognitive
radio system, the secondary users sense the channel independently. Then, the secondary users report to
the data collector where the OR rule is applied to determine the final spectrum sensing result. A method
based on the double-sided neighbor distance algorithm is proposed to detect attacks. A T -dimensional
space data, where T is the number of spectrum sensing periods, is used for detection. The distance (e.g.,
Euclidean, Hamming, and double-sided neighbor) is calculated from this historical sensing data, which
is used to identify the outlier secondary user far away from most secondary users in the history space.
Performance evaluation is performed to analyze the probability that the malicious user sends the sensing
result different from that of the secondary user.
In addition to the technical considerations, the design, optimization, and deployment of cognitive radio
networks will be affected by economic considerations. In particular, market mechanisms will be used
in designing efficient spectrum allocation and sharing methods. In the following section, we survey the
related work on different issues related to the economics of cognitive radio.
VI. ECONOMICS OF COGNITIVE RADIO NETWORKS
In a cognitive radio system, pricing is an important issue, which motivates the primary and secondary
users to share the available spectrum through a process often called spectrum trading [156]. Spectrum
trading is the mechanism for the entities in a cognitive radio system (e.g., primary and secondary users,
spectrum owner and users, service provider and subscriber) to exchange the radio resources. The exchange
could be performed through money or through different form of resources (i.e., bartering). In spectrum
trading, the primary users or service providers attempt to sell unused spectrum resources to secondary
users for monetary gains and the secondary users attempt to buy these resources to achieve their desired
communication goals. Two major approaches for spectrum trading are based on auction and open market.
A. Spectrum Auction
Auction is a traditional but efficient way to distribute commodity and services to the customers. The
theory of auction can be naturally applied to cognitive radio systems since the price of radio spectrum
cannot be determined precisely in advance [157]. In an auction, the primary users submit their asks and the
Page 32 of 66
http://mc.manuscriptcentral.com/wcm
Wireless Communications and Mobile Computing
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
For Peer Review
31
secondary users submit their bids to sell and buy a radio resource, respectively. Different auction models
have been applied to cognitive radio (e.g., single-sided, double-sided, English, Dutch, Vickrey-Clarke-
Groves (VCG) and combinatorial auctions). In [158], an auction-based algorithm (second-price auction
mechanism) is developed, which helps the wireless users to compete for channels in a fair manner. In this
auction, the users bid for a channel given the channel state. The users can optimize their bids to maximize
the throughput under the budget constraint. Game theory is applied to analyze the bidding strategies of
users and it is shown that the Nash equilibrium exists which also leads to the unique allocation of channel
under certain channel state distribution. In [159], different auction models (i.e., single object pay-as-bid
ascending clock auction [ACA-S], traditional ascending clock auction [ACA-T], and alternative ascending
clock auction [ACA-A]) are proposed to sell the spectrum of a primary user to multiple secondary users.
A multimedia streaming application with delay requirement is considered for secondary users to bid for
the spectrum. All of the proposed auction algorithms (i.e., ACA-S, ACA-T, and ACA-A) are distributed
and all of them converge within a finite number of auction rounds.
[160] considers spectrum auction in which a regulator sells channels for primary and secondary
networks. The primary network on each channel has higher priority to transmit data than that of secondary
networks, while all secondary networks have the same priority. Two objectives are considered, i.e., to
maximize revenue of regulator (i.e., auctioneer) and to maximize social welfare (i.e., sum of the valuations
of all networks). Different algorithms are developed for these objectives. It is also shown that the optimal
channel access allocation algorithm of dependent bids (i.e., the network has some knowledge or belief about
transmission pattern of other networks) is an NP-complete problem. Then, for the case of independent bids,
the dynamic programming algorithm is developed to solve the channel access allocation problem. [161]
studies the truthfulness of secondary users to inform the valuation and arrival-departure time period in
the spectrum auction performed by the primary user. The spectrum has the expiry constraint, where the
spectrum is only available for a certain period of time. In this auction, the secondary users have an
incentive to misreport the valuation and arrival-departure time period to gain more spectrum allocation
from the primary users. Therefore, to avoid such misreporting, the channel allocation and pricing schemes
are introduced to minimize the impact of false information.
To adapt the bidding strategy competitively, the bidders (e.g., secondary users) must have knowledge of
other opponents. However, information about other bidders may not be available publicly and the bidders
must learn from past auction results and history. In a cognitive radio, this learning capability can be
implemented for auction [162], [163]. In [162], a learning algorithm is adopted to optimize the repeated
Page 33 of 66
http://mc.manuscriptcentral.com/wcm
Wireless Communications and Mobile Computing
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
For Peer Review
32
spectrum auction in a cognitive radio network. A repeated spectrum auction is used which considers the
monitoring and entry costs of the secondary users to bid for the radio resource from the primary users.
Since in a distributed environment the knowledge of the secondary users about other users is limited, the
secondary users learn from experience by a Bayesian nonparametric belief update scheme and adapt the
bidding strategies accordingly. Based on the belief, the secondary user decides to join the auction or not.
In addition, the bidding strategy is proved to be optimal.
B. Price Competition in Open Market
Unlike an auction, which needs an auctioneer to control the trading, in an open market model, the
primary users and secondary users are free to sell and buy radio resources, respectively. Since there
is no control, the pricing strategy of primary users plays an important role here which determines the
revenue generated by the primary users. Also, pricing influences the decision of the secondary users to
buy the radio resource. A competitive pricing scheme based on a non-cooperative game among multiple
primary networks is proposed in [164]. A static game model and a dynamic game model are introduced
with and without global information, respectively. In the dynamic game model without complete global
information, the primary networks can adjust their pricing strategies to reach the Nash equilibrium. In
addition to non-cooperative pricing, the collusion among the primary users is analyzed, in which if the
collusion happens, the primary networks will adopt a cooperative pricing scheme instead of a competitive
one.
In [165], a joint power/channel allocation scheme is presented based on a distributed pricing approach.
The spectrum allocation is modeled as a non-cooperative game where the players are cognitive radio
users. A price-based iterative water-filling algorithm is introduced to reach the Nash equilibrium solution.
The game theory can model the behavior of cognitive radio users with self-interest, and therefore,
the game theoretic-models are very useful in modeling and optimization of dynamic spectrum sharing
networks [166]. With game-theoretic models, it is possible to design dynamic spectrum access methods
which can achieve flexibility, efficiency, and fairness. The work in [167] adopts a dynamic game model
to analyze competitive spectrum sharing. The importance of the work in [167] lies in the consideration
of dynamic behavior of strategy adaptation of primary users to sell the radio spectrum. In this case,
the convergence and the optimality of strategy adaptation are evaluated. The conditions to ensure the
stability of strategy adaptation are derived which will be useful for the primary users to avoid undesirable
fluctuations.
Page 34 of 66
http://mc.manuscriptcentral.com/wcm
Wireless Communications and Mobile Computing
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
For Peer Review
33
Primary user 1
Spectrum F1 Spectrum F2 Spectrum FM
Primary user 2 Primary user M
A group ofsecondary user 1
Group AGroup 2
Fig. 3: Spectrum trading in cognitive radio with multiple spectrum sellers and buyers [168].
In the most general scenario, spectrum trading in a cognitive radio system may involve multiple spectrum
sellers and buyers. In [168], the authors develop a spectrum trading framework for this general spectrum
trading scenario (Fig. 3). An evolutionary game model is used for the spectrum selection of secondary
users when choosing the primary user (or service providers) to buy spectrum from. A hierarchical game
model is presented to obtain the equilibrium solution for the primary users in selecting the spectrum price
to maximize their profit given the performance degradation of the primary users due to sharing available
spectrum with the secondary users. The collusive behavior of users could be a significant threat to efficient
dynamic spectrum allocation in a distributed cognitive radio network. [169] presents a systematic approach
to avoid collusion in cognitive radio. In [169], the spectrum allocation with multiple selfish legacy spectrum
holders and unlicensed users is modeled as a multi-stage dynamic game, and a pricing-based distributed
collusion-resistant spectrum allocation approach is used to optimize overall spectrum efficiency. [170]
studies price-based spectrum access control in which the primary users sell spectrum to secondary users.
The secondary compete with each other through random access. Therefore, the concept of “price of
contention” is introduced. The decentralized pricing mechanisms are developed for the monopoly and
multiple primary user market. The solution in terms of a Nash equilibrium is considered. [171] considers
spectrum leasing problem between primary and secondary users. Noncooperative game is formulated to
determine the optimal parameters.
For the deployment of practical cognitive radio systems, development of simulation tools, testbeds, and
hardware prototypes are essential. The simulation tools provide the capability for the hardware, software,
and protocol developers, engineers, and researchers to test and evaluate the design and new concepts for
cognitive radios with low cost, small effort, and short time. In addition, the testbed and the hardware
platforms can be used to develop prototypes for investigation and experimentation of the actual cognitive
Page 35 of 66
http://mc.manuscriptcentral.com/wcm
Wireless Communications and Mobile Computing
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
For Peer Review
34
radio systems. In the next section, we review the simulation tools, testbeds, and hardware prototypes
developed for cognitive radio systems.
VII. COGNITIVE RADIO SIMULATION TOOLS, TESTBEDS AND HARDWARE PROTOTYPES
A few work in the literature have introduced and reported simulation tools, testbeds, and hardware
prototypes for cognitive radio. [172] describes a simulation tool for performance evaluation and assessment
of a cognitive radio network with existence of multiple wireless systems (i.e., cellular networks and
opportunistic networks which are primary and secondary users, respectively). The simulation tool considers
the high-speed downlink packet access (HSDPA), hybrid ARQ (HARQ), and adaptive modulation and
coding (AMC) features for cognitive radios. [173] presents a cognitive radio simulation platform. The
customizable and extensible platform is developed based on C/C++ language. The major components in
the platform are the network topology, channel access of a primary user, mobility model, protocols (e.g.,
MAC and routing), traffic scheduling, and physical wireless channel. The platform has a set of performance
measures to be obtained, including throughput, spectrum efficiency, packet loss rate, jitter and delay.
[174] considers the issue of random number generation used for frequency-dependent spatial correlation
and position applicable for cognitive radio simulations. It is argued that the Kronecker approximation
cannot account for the dependencies. Therefore, a new method is introduced which uses nonassociative
left and right products of user-dependent frequency correlation matrices and frequency-dependent spatial
correlation matrices. While [174] concerns only random number generator, [173] provides a complete set
of simulation modules for cognitive radio.
Analog-digitalConverter (ADC)
Digitalprocessing
Digital-analogconverter (DAC)
Power amplifier
Isolator
Converter (ADC)
Fig. 4: Software-defined radio introduced by Mitola [176], [177].
[175] provides a comprehensive survey on the design methodologies for constructing flexible software-
defined radio receiver for cognitive radio based on the concept proposed by Mitola [176], [177] (Fig. 4).
The receiver should be able to receive data on a channel with any band from 800 MHz to 6 GHz and
any bandwidth. The design of the receiver is proposed to have a digital front-end which can be tuned
electronically. Then, the analog baseband samples the channel of interest at zero intermediate frequency
Page 36 of 66
http://mc.manuscriptcentral.com/wcm
Wireless Communications and Mobile Computing
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
For Peer Review
35
(IF). The clock-programmable downsampling is applied with the embedded filtering. With this design, the
receiver can change the operating frequency band dynamically. The use of this design in CDMA requires
the system to operate in full-duplex mode (i.e., the transmitter and the receiver operate simultaneously),
and the ability of the receiver to operate on multiple channels concurrently.
[178] introduces a real-time cognitive radio testbed for physical and link layer experiments. This testbed
is based on the Berkeley Emulation Engine 2 (BEE2), which is a multi-FPGA emulation platform. Using
BEE2, the testbed can connect to 18 radio front-ends, and these front-ends can be set to be the primary
and secondary users in various test scenarios. In addition, by using FPGA, the testbed can simultaneously
operate on multiple radios, which is suitable for complex experiments and performance studies involving
high-speed low latency links. [179] discusses the design and implementation of an UWB cognitive radio for
high data rate communications with very low transmission power. The design of ultra-low power baseband,
impulse-UWB transceiver front-end, and digital backend implementation is introduced. In addition, [179]
reports a testbed implementation based on the BEE2. In [180], the design of a software-defined radio from
a low-power ADC perspective is introduced. The design is based on the windowed integration sampler
and clock-programmable discrete-time analog filters. By using a low-noise amplifier (LNA) and a wide
tuning-range synthesizer, the wideband RF front-end can operate on any frequency between 800 MHz
and 6 GHz. The wideband LNA can achieve a good performance (e.g., 18-20 dB of maximum gain). The
performance evaluation is performed on the GSM and the IEEE 802.11g standards.
There are a number of wireless testbeds with cognitive networking capabilities which have been
developed in the academia and they differ in terms of the scale of the network, the degrees of experimental
control and isolation they offer, and the indoor and/or outdoor nature of the deployment and measurement
environment [3]. They are based on the off-the-shelf radio technologies such as Wi-Fi and WiMAX
technologies. A variety of software radio platforms for base stations, client nodes, spectrum sensors,
and different RF sources or users are used in these testbeds. These platforms are generally capable of
implementing different modulation algorithms, MAC protocols, and adaptation methods. Examples of
the testbeds include the CMU emulator [181], the ORBIT radio grid testbed [182], the “DieselNet”
testbed [183], and the “CitySense” network [184]. The CMU wireless network emulator testbed has 15
nodes (with 802.11b or Bluetooth interfaces) and it supports the entire 2.4 GHz ISM band. The nodes
communicate through the emulator, which is an FGGA-based central DSP engine, and this emulator
models the effects of wireless signal propagation. The ORBIT radio grid testbed, which contains 400
programmable radio nodes in an area of 5000 sq-ft, is also based on an emulator, and the testbed can
Page 37 of 66
http://mc.manuscriptcentral.com/wcm
Wireless Communications and Mobile Computing
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
For Peer Review
36
be accessed through an Internet portal. It supports end-to-end wired and wireless experiments. Also, the
testbed includes an outdoor field trial system. Examples of cognitive radio platforms include WiNC2R
cognitive radio board [185], the Kansas University Agile Radio (KUAR) [186], the Defense Advanced
Research Projects Agency (DARPA) Wireless Network after Next (WNaN) cognitive radio [187], the
Universal Software Radio Peripheral (USRP) and GNU radio software architecture [188], and the RICE
WARP board [189]. In general, these platforms consist of RF transceiver modules, digital processing
boards, and networking interfaces/modules, and they are supported by software architectures such as
GNU.
VIII. FUTURE TRENDS IN COGNITIVE RADIO NETWORKING RESEARCH AND OPEN ISSUES
In this section, we summarize future trends in cognitive radio networking research which include
the cooperative and cognitive radio communications, cognitive radios for the emerging LTE-Advanced
networks, and cognitive radio applications. Also, the major open research issues are outlined.
A. Cooperative and Cognitive Communications
Relay-based communication is a powerful approach to combat channel fading, extend transmission
coverage, and improve system capacity [190]. This approach can be also used for cognitive radio net-
works to improve the network capacity, reduce interference to primary users, and achieve fairness in
data transmission among the cognitive radio users [191]–[196]. However, due to the different spectrum
availability at the cognitive radio nodes (i.e., transmitter, receiver, and relay) in a cognitive relay network,
the traditional relay-based communication approaches cannot be directly applied.
The relay-assisted cognitive transmission can be divided into two main scenarios, i.e., cooperative
transmission between secondary users [197], [198] and cooperative transmission between primary users and
secondary users [191], [199]. In the first scenario, the secondary users act as a relay to assist only secondary
user’s transmissions. In the second scenario, each secondary relay cooperates by relaying primary user’s
signal to its primary receiver, and as an incentive, each secondary relay gets an opportunity to transmit
in the same spectrum which is used by the primary user. One of the most difficult challenges in relay-
based cognitive systems is the design of MAC protocols which are able to utilize the best relay and
the spectrum opportunities for the optimal performance. A centralized relay assisted MAC protocol is
presented in [191], [194] for relay-assisted cognitive networks.
Page 38 of 66
http://mc.manuscriptcentral.com/wcm
Wireless Communications and Mobile Computing
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
For Peer Review
37
For resource allocation in relay-based cognitive radio networks, in addition to power and rate allocation,
the channel and relay allocation are also important since the secondary users as well as the relays contend
for channel access. Thus, the allocation of both relay and channel resource is coupled together and must
be addressed jointly [197], [198], [200]–[205]. The selective fusion spectrum sensing and best relay data
transmission (SFSS-BRDT) scheme and the fixed fusion spectrum sensing and best relay data transmission
(FFSS-BRDT) scheme are used to address this relay selection problem [206]. In [207]–[209], centralized
solutions are considered in which the resource allocation at the transmitter (cognitive radio base station)
and at the relay stations is computed in a centralized manner based on the global channel state knowledge.
Due to the signaling overhead and computational complexity, decentralized resource allocation is more
desirable for cognitive relay networks [210]. In practical implementation, due to imperfect channel sensing,
the relay node may not always be able to acquire available spectrum for relay transmission. Further, due
to the imperfect channel state information (CSI), the selected relay may not always be the best relay. The
problem of optimal resource allocation under imperfect spectrum sensing and imperfect CSI in cognitive
relay networks has been addressed in the recent literature [199], [210]–[212].
B. Cognitive Radio for LTE/LTE-Advanced Networks
Dynamic resource management is the one of the most significant features in cognitive and software-
defined radio systems. Future LTE/LTE-Advanced networks will have to address more complex resource
management problems involving both contiguous and non-contiguous carrier aggregation, dynamic co-
channel and adjacent channel interference mitigation, advanced multiple-input multiple-output (MIMO)
techniques, multiple radio access technology (RAT) functionalities, coordinated multiple point (CoMP)
transmission and reception, and relaying for efficient use of radio spectrum [213]. Cognitive radio concepts
will be useful to develop enabling techniques to implement these features in the LTE/LTE-Advanced net-
works. LTE-Advanced networks will also use the concept of femtocells (also referred to as heterogeneous
networks) to provide higher quality of service to the end users by improving the indoor coverage. However,
deployment of femtocells will result in a complex interference scenario when compared to that due to
traditional frequency planning. Interference mitigation in such a scenario will require frequency agility,
dynamic channel selection, and complex power control. The cognitive radio technologies can be used in
the heterogeneous networks to enable the femtocells to perform the required adaptation.
Page 39 of 66
http://mc.manuscriptcentral.com/wcm
Wireless Communications and Mobile Computing
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
For Peer Review
38
C. Cognitive Radio Applications
1) General applications: A comprehensive survey of the emerging cognitive radio applications is
presented in [214]. This paper identifies some killer applications of cognitive radio technology and
discusses about the opportunities. The major applications include smart grid communications, public
safety communications, and broadband cellular to medical applications. This paper discusses how to
adopt cognitive radio in these applications. The benefits and the challenges are outlined and the possible
solutions are discussed. Last but not the least, this paper reviews some standardization activities related
to adopting cognitive radio in various applications.
The potential applications of cognitive radio to the public safety communications are discussed in [215].
The public-safety community will adopt dynamic spectrum access technique to improve the efficiency,
capacity, robustness, and flexibility of the wireless communications systems. This paper raises an important
issue regarding such an adoption which is the spectrum management reform to create spectrum pools.
With these spectrum pools, cognitive radio devices can utilize the spectrum efficiently under the control of
adaptive prioritization policies. This approach overcomes many limitations in the legacy communications
systems. This paper also discusses about the economic, policy, and business issues of such spectrum
pooling.
The work in [216] discusses the use of cognitive radio in the IEEE 802.16-based WiMAX networks.
Although the original IEEE 802.16 standard does not adopt any cognitive radio into account, this paper
points out that significant benefit can be achieved by increasing frequency reuse through power control and
cognitive channel assignment. Then, the paper introduces an algorithm for optimal channel assignment.
Simulations are performed for an IEEE 802.16 system to verify the benefits of the proposed algorithm.
One of the very promising applications of cognitive radio will be its use in the evolving femtocell
networks to improve the indoor coverage as well as the capacity of traditional cellular wireless networks.
The work in [217] introduces an interesting idea of ultra-broadband femtocells based on the cognitive radio
concept. Specifically, the opportunistic spectrum reuse method is presented which allows the femtocells
to access the radio spectrum licensed to different operators with different services. The architecture
and the key technology enablers are discussed to achieve the optimal performances of heterogeneous
networks with multiple macrocells and femtocells. The work in [218] proposes a femtocell-based cognitive
radio architecture to enable multitiered opportunistic access in next-generation heterogeneous networks
(HetNets). The architecture unifies the concept of conventional femtocells with infrastructure-based overlay
Page 40 of 66
http://mc.manuscriptcentral.com/wcm
Wireless Communications and Mobile Computing
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
For Peer Review
39
cognitive radio networks.
The principles of cognitive radio can be also used in multihop wireless mesh networks. Instead of a
fixed channel allocation among the mesh nodes, a dynamic channel allocation can be used to improve
the spectrum utilization. With this spirit, in [219], a cluster-based dynamic channel allocation framework
is proposed for cognitive wireless mesh networks taking the issues of interference and coexistence with
primary users into account. In [220], the authors present applications of cognitive radio in the context
of green communications. Two approaches are discussed, i.e., applying cognitive radio to achieve energy
efficiency and developing energy efficiency for cognitive radio systems. The paper introduces different
techniques for achieving energy efficiency in wireless systems and how to apply those techniques in
cognitive radio.
2) Cognitive radio networking in the TV white space: Cognitive radio networks can be developed to
exploit the TV white space (TVWS) for broadband wireless access, as the analog TV bands, which is
largely underutilized, has been released for data communications. Many studies in the literature consider
cognitive radio on the TV white space (e.g., [221]–[224]). In the TVWS, the TV broadcasting stations
and low-power wireless microphones serve as incumbents and the secondary systems such as the IEEE
802.22-based wireless regional area networks, the Wi-Fi hotspots and home networks can coexist with
the incumbent systems.
[221] provides a comprehensive survey on accessing TVWS. Various issues are identified, including
high-precision spectrum sensing [225], agile modulation and spectrum pooling techniques, and system-
level design. Also, the potential of using TVWS for a wireless home network and cognitive femtocell is
discussed. The coexistence challenges for heterogeneous cognitive networks in the TVWS were discussed
in [222].
In [223], the problem of coexistence among wireless networks in TV white space is studied. The
framework to exchange information related to coexistence is also introduced. This framework supports
both centralized and distributed information exchanges. The framework uses a multiradio cluster-head
equipment (CHE) to collect information about TV white space and coexisting wireless networks. In the
centralized approach, there is a central coexistence database available for the CHE to obtain information
about wireless environment and to make the decision on spectrum access. On the other hand, in the
distributed approach without the central database, a broadcast channel is used to exchange the coexistence
information among CHEs. The simulation results show that with the proposed framework, both centralized
and distributed approaches can discover larger spectral availability than that without information exchange.
Page 41 of 66
http://mc.manuscriptcentral.com/wcm
Wireless Communications and Mobile Computing
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
For Peer Review
40
In [226], a wireless hardware prototype to operate on TV white space is introduced. The prototype has
a sensor and a geolocation device to collect information about the wireless environment in TV bands. This
information is stored in the TV band database. A cognitive management entity is introduced to process
the channel information and to control the transmission and reception of the transmitter and receiver,
respectively. The prototype is developed based on the TV white space test trial regulation in Singapore
which requires the wireless networks to operate on the 630-742 MHz band. The prototype is capable to
automatically search for and access unused TV bands.
[227] presents the feasibility study of using TV white space for wireless home networks. Also, the
performance is compared with that of 2.4 GHz and 5 GHz ISM bands which are currently widely used.
The simulation results show that the performance of using TV white space is better than that of 5 GHz
and 2.4 GHz bands for low to medium traffic load (e.g., several Mbps). Also, low power level (i.e., below
3 dBm per channel) can be used. However, at high traffic load, it is suggested that the TV bands should
be used with the interference management, and are suitable as the complementary channels of the 5 GHz
and 2.4 GHz bands.
3) Cognitive radio for smart grid communications: The concept of smart grid has been introduced
to improve the efficiency and responsiveness of the traditional power grid. The smart grid will use data
communications technologies to transfer different information related to the generation, transmission, and
distribution operations of the power grid to achieve the optimal operation of the power system. Wireless
technologies will be used for the smart grid communications [228]. Cognitive radio will be a potential
technology to improve the spectrum utilization and transmission efficiency of wireless communications
used in the smart grid [229]–[231].
In [229], the feasibility of using cognitive radio in the smart grid applications is extensively investigated.
The system architecture, algorithms, and hardware testbed are developed and analyzed for a microgrid
environment to support power and information flows. In addition, the mechanisms to ensure secure control
and information exchange in the power system are also considered. The independent component analysis
(ICA) together with the robust principal component analysis (PCA) techniques are used to improve the
robustness of the smart meter data transmission under strong interference. In [232], a hierarchical cognitive
radio communications infrastructure is proposed as a solution to manage a large number of transceivers
in the smart grid. In this infrastructure, cognitive radio will be potentially deployed for the home area
networks (HANs), the neighborhood area networks (NANs), and the wide area networks (WANs). Dynamic
spectrum access and sharing is also proposed to optimize the performance in each network to achieve the
Page 42 of 66
http://mc.manuscriptcentral.com/wcm
Wireless Communications and Mobile Computing
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
For Peer Review
41
globally optimal spectrum access strategy. In [233], a cognitive radio network is developed for HANs.
The spectrum sensing and channel switching techniques are introduced for the smart meter to transmit
the meter data on either the licensed or the unlicensed channels. The objective is to improve reliability
and reduce delay in data transmission. The sensing time is optimized to reduce packet loss rate and delay
while the transmission of primary user is not interfered.
Although the above work addresses different issues of using cognitive radio in the smart grid, there are
many research opportunities. For example, a QoS framework for the cognitive radio used in smart grid
must be designed and optimized to meet the delay and reliability requirements of the different smart grid
applications. The benefits of cognitive radio in power generation, power transmission, and consumption
must be analyzed. Also, the integration of the cognitive radio into existing wireless communications
infrastructure of the smart grid needs to be investigated.
4) Cognitive radio for machine-to-machine (M2M) communications: Machine-to-machine (M2M) com-
munications allows wireless nodes to communicate without or minimal interaction with human. In M2M
communications, the data to be transmitted is smaller in size but more in frequency than those of the
human-to-human (H2H) communications. Cognitive radio is a potential solution for M2M communications
by offloading the data transmission from licensed channel to unlicensed channel. However, there are
many issues that need to be addressed. M2M communications will be application centric, in which the
protocol design and optimization must take the requirements of the specific applications into account. For
example, [234] considers M2M communications for the smart grid applications focusing on the energy
consumption of data transmission through cognitive radio networks (i.e., TV band devices [TVBDs]). The
dynamic spectrum access is optimized using a multi-objective genetic algorithm to find the optimal tradeoff
between power efficiency and spectrum efficiency. Also, the spectrum allocation for M2M communica-
tions must be determined. [235] studies the cognitive radio network tomography and spectrum map for
cooperative spectrum sensing. A two-step cooperative learning algorithm is proposed to analyze the spatial
correlation of the results from sensing nodes and the spectrum map is efficiently constructed. The proposed
algorithm is suitable for large-scale, dynamic, and heterogeneous M2M networks. M2M communications
can be integrated into the existing wireless systems (e.g., WiMAX and LTE [236] networks). However,
optimal system architectures must be developed for integrating M2M communications and cognitive radio
into such wireless systems. With the dense deployment of D2D communications in the limited cellular
radio spectrum, cognitive radio-based sophisticated interference control and avoidance methods will be
essential for these systems.
Page 43 of 66
http://mc.manuscriptcentral.com/wcm
Wireless Communications and Mobile Computing
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
For Peer Review
42
5) Cognitive radio and mobile cloud computing: Mobile cloud computing combines wireless technolo-
gies and cloud computing to improve the performance and reduce cost of mobile services to the users [237].
In mobile cloud computing, the mobile application is divided into two parts, i.e., local computing modules
to run on a mobile device and remote computing modules to run on a server in the cloud to reduce the
energy consumption of a mobile device and to improve the performance of a mobile application. These
two parts need to communicate with each other through a wireless connection. Cognitive radio can be
used to support mobile cloud computing, i.e., to transfer data among local and remote modules of mobile
applications. The major challenge in using cognitive radio for mobile cloud computing is at the problem of
guaranteeing QoS to the users and efficient radio resource allocation [238]. The radio resource allocation
must be performed jointly with the computing resource (e.g., CPU of a server) allocation since mobile
applications can be executed in different modes, each of which requires different amount of bandwidth
and computing resource. Energy efficiency will be also an important factor which has to be optimized for
cognitive radios used in a mobile cloud computing scenario.
Cloud computing can be used to support cognitive ratio networks. Complex signal processing for
wireless transmission can be offloaded from mobile devices and base stations to the servers in the cloud.
This concept has been used in a few work in the literature. In [239], cloud computing is used as the
central communication and optimization infrastructure for a cognitive radio network. This network is
used for advanced metering infrastructure (AMI) to support the smart grid applications. In this case, the
data processing of AMI is performed by servers in the cloud. In [240], cooperative spectrum sensing
and radio resource scheduling for cognitive radio communications in the TV white space is implemented
and executed in a cloud computing environment. In particular, the spectrum sensing reports are analyzed
in the cognitive radio cloud (CRC) to identify the spectrum opportunities. The analysis is done through
the sparse Bayesian learning algorithm using Microsoft’s Windows Azure Cloud platform, which is one
of the public cloud computing services. In [241], the authors present a prototype system of a cognitive
radio architecture implemented using cloud computing to store and process spectrum sensing data. Cloud
computing is used for data sharing among network devices in the network centric operation.
D. Open Issues
The emerging cognitive radio technology is envisioned to become a key element in the future generation
“smart radio” technology which will dynamically optimize usage of limited radio resources and mitigate
the “radio traffic jam” problem in future wireless networks. This technology will also provide robustness
Page 44 of 66
http://mc.manuscriptcentral.com/wcm
Wireless Communications and Mobile Computing
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
For Peer Review
43
and efficiency for wireless and mobile communications services. It is anticipated that cognitive radio will
evolve as a general-purpose programmable radio, which can be implemented through SDR, and it will serve
as a universal platform for flexible and adaptive wireless system developments [3]. Although significant
amount of research efforts have been spent on different aspects of cognitive radio communications and
networking, there are still many technical challenges and interesting research opportunities. Also, there
are significant challenges in the spectrum policy domain and pricing for dynamic spectrum access. In
addition, rigorous testing and evaluation of cognitive radio technology in real-world environments is a
significant issue which must be addressed to study the emergent behavior of cognitive radio and for
large-scale deployment of this technology. A few of the open technical challenges are outlined below.
1) Practical methods to estimate channel parameters and spectrum sensing for cognitive radios: A
cognitive radio system is characterized by a time-dependent structure of the primary user traffic pattern and
imperfect spectrum sensing. It has been investigated by a large body of cognitive radio literature, especially
from the standpoint of the decision-theoretic framework such as a partially observable Markov decision
process (POMDP), quickest detection, sequential detection, a restless bandit problem, and sequence
detection. In most of the existing work in the literature, secondary users’ knowledge of the channel
parameters is simply assumed without giving a specific method for obtaining the information. However,
in a real scenario, it may be the case that the primary user does not explicitly help the secondary user
in obtaining the knowledge of the channel parameters. Therefore, developing a method for the secondary
users to estimate the channel parameters without any help from the primary user is crucial for the design of
a practical cognitive radio system. Although a significant amount of research have been done on spectrum
sensing based on transmitter detection, there are only a limited number of techniques available for receiver
detection.
2) Practical spectrum sharing protocols for cognitive radios: Practical methods for spectrum sharing
among cognitive radios need to be developed considering the limited (or no) availability of control
channels for coordination of spectrum access in a multi-user, multi-channel, and multi-radio scenario.
Both horizontal and vertical spectrum sharing scenarios need to be considered. In this context, distributed
adaptation and resource management techniques, which depend on local information only and incur low
control overhead and implementation complexity, are desirable. For this, different machine learning, pattern
recognition, and artificial intelligence tools will be useful. In order to make the developed resource
allocation (i.e., power and rate allocation) methods amenable to real-world environments, they have
to be made robust against the various limitations encountered in practice such as the unavailability
Page 45 of 66
http://mc.manuscriptcentral.com/wcm
Wireless Communications and Mobile Computing
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
For Peer Review
44
of instantaneous channel gain information (due to the time-varying nature of wireless channel) and
unavailability (or limited availability) of a control channel.
Also, practical protocol designs to support spectrum sharing using advanced communications techniques
such as network coding, multiple-input multiple-output (MIMO) communication, relay-based cooperative
communication in cognitive radio networks are interesting research problems.
Since the cognitive radio systems could be vulnerable due to the malicious eavesdroppers who try to
overhear other users’ messages. Therefore, spectrum access and sharing methods are required which can
maximize the secrecy rate (i.e., rate of perfectly secure transmission of information) of cognitive radio
users against the malicious behavior of eavesdroppers.
3) Protocol architectures and cross-layer adaptation: Design and evaluation of higher layer protocols
such as the transport layer protocols for cognitive radio networks in a wired-cum-wireless scenario have
not been adequately addressed in the literature. Design and analysis of transport and routing layer protocols
for a multi-hop or ad hoc cognitive radio network [242] also need extensive investigation. Cross-layer
adaptation and optimization will be a key component in the design of cognitive radio protocols for adaptive,
flexible, and stable dynamic spectrum access networks (Fig. 1). This optimization will need to consider
spectrum availability and dynamic channel and network selection in addition to the network parameters
such as QoS requirements, mobility, traffic load, node density, channel parameters as considered for
traditional cross-layer optimization in wireless networks. Also, a management and control framework will
be required for exchange of inter-layer information to support cognition and achieve cross-layer adaptation.
Design and evaluation of such frameworks is an open problem.
4) Spectrum policy research on cognitive radio: Spectrum policy research is required to define the
etiquettes for the cognitive radios, analyze their impacts on the technology and business strategies, analyze
their economic and social benefits, develop mechanisms to enforce the etiquettes derived from the spectrum
policy, and define the role of the different stakeholders [3]. Development of policy and legal frameworks
for cognitive radios for the different spectrum sharing models (or market models) is a very significant
research direction. Techniques from microeconomic and game theory models and incentive mechanisms
are useful in this context. Although there have been some work in this direction, more work are required to
analyze large-scale spectrum markets in terms of efficiency and convergence and understand their emergent
behavior as well as practical implementation of the derived methods and policies in the cognitive radio
protocols. Enforcement of the spectrum policies for the cognitive radios is another big challenge.
Page 46 of 66
http://mc.manuscriptcentral.com/wcm
Wireless Communications and Mobile Computing
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
For Peer Review
45
5) Development of cognitive radio testbed, experimentation, and application-centric performance eval-
uation: Since cognitive radio networks are very complex systems with many degrees of freedom, it is very
critical to evaluate their system-wide performance in real-world environments and study their emergent
behavior to assess the viability of their practical implementation. Therefore, development of prototypes and
testbeds and their management are essential so that experimentations can be done in realistic environments
to complement the analytical models and simulations. The adaptive protocols and the policy requirements
add challenges for the testbed development and experimentation for cognitive radio networks. Again,
application-centric evaluations (e.g., for public safety applications, broadband Internet access in rural
areas, smart grid communications, eHealth applications, vehicular networking applications) of cognitive
radio systems are interesting avenues of future research on cognitive radio.
6) Cognitive small cell networks: Deployment of small cells has been identified as a potential approach
to increase the cellular wireless network capacity. By connecting to a small cell base station (e.g., a pico
or femto base station for picocell and femtocell, respectively), users can locally transmit and receive data
with much better channel quality. However, the small cells will be underlaid in a macrocell. If the small
cells and the macrocell use the same spectrum, interference will occur, deteriorating the performances of
users in both cell tiers. Cognitive radio techniques can be adopted to mitigate this problem. In this case,
the small cells can reuse the spectrum allocated to the macrocell. By opportunistic access, the spectrum
utilization can be improved while the network capacity is enhanced due to macrocell offloading [243].
For efficient opportunistic spectrum access, spectrum sensing mechanism to identify spectrum availability
needs to be carefully and optimally designed [244].
7) Cognitive Device-to-Device (D2D) communications: D2D communications has been introduced to
offload data traffic from a base station by allowing D2D devices to communicate with each other directly.
Similar to cognitive small cell networks, D2D communications can reuse the spectrum allocated to the
cellular users (i.e., users connected with the base station). Cognitive radio techniques can be applied to
D2D devices (i.e., secondary users) to opportunistically access the spectrum. Again, the spectrum sensing
is important to ensure that QoS requirements of the cellular users will not be violated while D2D devices
can achieve satisfactory performance.
8) Energy-efficient cognitive radio: Energy is an important resource for mobile devices [245]. There-
fore, techniques in cognitive radios (e.g., spectrum sensing [246], dynamic spectrum access [247] [248] and
sharing [249]) can be also optimized to maximize energy efficiency for data transmission. One approach for
spectrum sensing is to use an adaptive algorithm to perform sophisticated signal detection and estimation,
Page 47 of 66
http://mc.manuscriptcentral.com/wcm
Wireless Communications and Mobile Computing
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
For Peer Review
46
which consumes a lot of energy, only when needed (e.g., a primary user is likely to present). For the
spectrum access, a secondary user must optimize the data transmission given the available energy [250].
Different forms of learning and reasoning techniques can be exploited for this. In the context of both single-
tier and multi-tier cellular wireless networks, cognitive radio techniques (e.g., for traffic load sensing and
estimation) can be used to opportunistically switch on/off some of the base stations while satisfying the
users’ communications requirements. Also, cognitive radio users can use energy harvesting techniques to
avoid relying on a fixed source of energy [251].
IX. COGNITIVE RADIO STANDARDIZATION
The IEEE 802.22 and SCC 41 are considered as the primary cognitive radio standards today. They are
also the completed standards of interest for cognitive radio [252]. However, there are also many other
standards being developed. IEEE created the 802.22 Working Group (WG) for WRANs in November,
2004. This WG has been assigned to develop an air interface (i.e., PHY and MAC) based on CRs for
unlicensed operation in the TV broadcast bands. The focus of 802.22 is on rural broadband wireless access
and its coverage range is considerably larger than that of the IEEE 802.16 [253]. A brief review of the
different standardization efforts is provided below.
A. IEEE 802.22
An overview of the 802.22 architecture (e.g., topology, entities, and connections), its requirements (e.g.,
service capacity, service coverage, physical and MAC layer details), applications, and coexistence issues
(e.g., antenna, TV and wireless microphone sensing and protection) are discussed in [254], [255]. In North
America, the frequency band of operation of the IEEE 802.22 networks is 54-862 MHz. Also, the standard
shall accommodate various international TV channel bandwidths of 6, 7, and 8 MHz. The IEEE 802.22
systems have a fixed point to multipoint air interface. The base station controls the consumer premise
equipments (CPEs). The base station also coordinates “distributed sensing” which is needed for incumbent
(original user) protection. The base station directs the CPEs to measure different TV channels and provide
feedback. For an average spectral efficiency of 3 bits/sec/Hz in the IEEE 802.22 systems, considering 12
users, the minimum peak throughput rate is 1.5 Mbps per CPE in the downstream direction and 384 kbps
in the upstream direction. Another striking feature of the IEEE 802.22 WRAN as compared to the existing
IEEE 802 standards is the base station coverage range, which can be up to 100 km if transmission power
is not an issue (current specified coverage range is 33 km at 4 watts CPE EIRP).
Page 48 of 66
http://mc.manuscriptcentral.com/wcm
Wireless Communications and Mobile Computing
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
For Peer Review
47
B. IEEE 1900 - SCC41- DYSPAN
The IEEE Standard Coordinating Committee 41 (SCC41), formerly known as the IEEE 1900 task
force [256], was created by the IEEE to work in the area of dynamic spectrum access networks for
cognitive radio standardization. The IEEE SCC41 is divided into four Working Groups (WGs) termed
as 1900.x, “x” being the WG. In 2010, the SCC41 was approached by IEEE Communications Society
Standards Board (CSSB), and was renamed as IEEE DYSPAN-SC.
IEEE P1900.1 (Terminology and Concepts for Next Generation Radio Systems and Spectrum Manage-
ment) intends to create a glossary of important cognitive radio terms and concepts related to spectrum
management, policy-defined radio, adaptive radio, software-defined radio, and related technology, and also
compare different technologies and their capabilities [257].
IEEE P1900.2 (Recommended Practice for Interference and Coexistence Analysis): The 1900.2 WG rec-
ommends interference analysis criteria and develops a structure for measuring/analyzing the interference.
This standard provides a systematic way of analyzing interference and coexistence.
IEEE P1900.3 (Dependability and Evaluation of Regulatory Compliance for Radio Systems with Dy-
namic Spectrum Access): On the software side, the 1900.3 WG is developing test methods for evaluating
SDR devices. The main goal is to verify the coexistence and compliance of the software modules for
cognitive radio devices before certifying final devices.
IEEE P1900.4 (Architectural Building Blocks Enabling Network-Device Distributed Decision Making
for Optimized Radio Resource Usage in Heterogeneous Wireless Access Networks): This standard is for
radio systems with multiple radio access technologies (RATs) [258], [259]. The end terminal users are
considered to be supporting multiple RATs and with some cognitive radio capabilities such as flexible
operations in different frequency bands. The IEEE 1900.4 defines reconfiguration management entities
(RMEs), which help in decision making at terminal and network sides.
IEEE P1900.5: The SCC41 established this working group to address the policy language issues.
IEEE P1900.6: The SCC41 established this working group to address RF sensing functions that can
be managed in a terminal reconfiguration manager (TRM) defined in the IEEE P1900.4. Published in
April 2011, this standard provides definition of structure and interfaces for exchanging sensing related
information between sensor and users of sensing information (cognitive engines).
IEEE P1900.7 (Radio Interface for White Space Dynamic Spectrum Access Radio Systems Supporting
Fixed and Mobile Operation):
The standard would focus on the development of cost-effective white space dynamic spectrum access
Page 49 of 66
http://mc.manuscriptcentral.com/wcm
Wireless Communications and Mobile Computing
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
For Peer Review
48
radio systems capable of operation in white space frequency bands on a non-interfering basis to incumbent
users in these bands.
The IEEE SCC41 has also started cooperating with the FCC, Federal office of Communication (OF-
COM) UK, SDR Forum and OMG forum, etc.
C. IEEE 802 Projects Related to Cognitive Technologies
IEEE 802.16.2: Completed in November 2001, it is one of the first coexistence standards which provides
guidelines for minimizing interference in fixed broadband wireless access (BWA) systems.
IEEE 802.15.2: Completed in June 2003, this standard outlines recommended practices for coexistence
of IEEE 802.15 wireless personal are networks (WPANs) with other wireless devices operating in unli-
censed bands. It also provides a technique for the IEEE 802.11 devices to coexist with the IEEE 802.15
devices.
IEEE 802.11h: Completed in September 2003, this standard provides mechanisms for dynamic frequency
selection (DFS) and transmit power control (TPC). The DFS feature developed for the IEEE 802.11 allows
relocation of interfering IEEE 802.11 basic service set (BSS) to another frequency. This standard has been
superseded by IEEE 802.11-2007.
IEEE 802.11y: This amendment to the IEEE 802.11 standard will allow 802.11 to operate at 3650-3700
MHz bands in the U.S. It will allow shared 802.11 operations with other users. It also includes transmit
power control (extending 802.11h) and DFS (extending 802.11h).
D. Wireless Innovation Forum (SDR Forum)
In 2008, the SDR forum published a report [260] on “Cognitive Radio Definitions and Nomenclature”.
This report highlights the components involved and defines many working technologies involved with
cognitive radio. The Cognitive Radio Working Group (CRWG) under SDR forum is also working on a
report entitled “Quantifying the Benefits of Cognitive Radio” developed for the benefits of using cognitive
radio technologies in upcoming wireless systems. In 2009, the SDR forum also started a test and measure-
ment project to develop a set of use cases, testing requirements, and methods for SSDR/CR technologies.
The project addresses critical TVWS functions, e.g., spectrum sensing, interference avoidance, database
performance, and adherence to policy.
Page 50 of 66
http://mc.manuscriptcentral.com/wcm
Wireless Communications and Mobile Computing
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
For Peer Review
49
E. ETSI Standardization
In October 2009, ETSIs Technical Committee (TC) on Reconfigurable Radio Systems (RRS) released
a series of ETSI Technical Reports (TR 102 838) [261] that examine the standardization needs and
opportunities related to reconfigurable radio systems including SDR and CR. The reports summarize the
feasibility studies carried out by the committee and present the recommended topics for standardization.
Following the completion of the feasibility studies, standardization of RRS is now getting underway with
four different working groups studying different aspects of reconfigurable radio systems as follows:
System Aspect Studies (WG1): The basic role of this group is to form a technical framework that can
map other topics being studied by other groups.
Reconfigurable Radio Equipment Studies (WG2): This group finalized two study reports identifying
potential SDR standardization Mobile Devices (MD) and Reconfigurable Base Stations (RBS).
Cognitive Network Management Studies (WG3): This study is divided into two parts, i.e., Functional
Architecture (FA) and Cognitive Pilot Channel (CPC). The scope of FA is to study and propose a generic
architecture for the management and control of reconfigurable radio systems, namely, the Functional
Architecture (FA). The FA is designed to improve the utilization of spectrum and radio resources.
Public Safety Studies (WG4): This group studies the application of RSS to the public safety domain
in two parts, i.e., user requirements and operational requirements.
F. International Telecommunication Union (ITU) Standardization
The ITU-Radio communication sector (ITU-R) study Group 8 (Mobile, Radio determination, Amateur
and Related Satellite Services) deals with standardization of cognitive radio networks. ITU-R study Group
8 has published two reports on SDR [262], [263]. The reports focus on the application of SDR in
International Mobile Telecommunications-2000 (IMT-2000) technology. IMT-2000 systems are the third
generation mobile systems, which provide access to a wide range of telecommunication services, supported
by the fixed telecommunication networks (e.g., PSTN/ISDN/IP), and to other services, which are specific
to mobile users. The SDR technology is employed in the base station and the controllers of a mobile
radio access network (RAN) to increase the flexibility of radio access networks.
X. CONCLUSION
Cognitive radio represents a new paradigm for designing intelligent wireless networks to mitigate
the spectrum scarcity problem and provide significant gain in spectrum efficiency. We have provided
Page 51 of 66
http://mc.manuscriptcentral.com/wcm
Wireless Communications and Mobile Computing
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
For Peer Review
50
a comprehensive survey of the research activities in cognitive radio. The major issues in the design of
cognitive radio communication networks have been discussed (e.g., spectrum sensing, dynamic spectrum
access, applications, and standardization) and the related work in the literature have been reviewed. First,
the historical note of the cognitive radio has been given to provide a motivation of the dynamic and efficient
next generation wireless systems. Then, the fundamental concepts and theoretical analyses of the dynamic
spectrum access have been presented. The various approaches of spectrum sharing in cognitive radio have
been reviewed. The security and economic issues have been discussed. From the practical perspective,
the reviews of simulation tools and testbed have been provided. The future trends and research directions
have been discussed. Also, the standardization activities related to cognitive radio have been summarized.
Page 52 of 66
http://mc.manuscriptcentral.com/wcm
Wireless Communications and Mobile Computing
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
For Peer Review
51
REFERENCES
[1] Guest Editorial, KICS/IEEE Journal of Communications and Networks (JCN) Special Issue on “Cognitive radio: A path in the evolution
of public wireless networks”, December 2008.
[2] S. Haykin, “Cognitive radio: Brain-empowered wireless communications,” IEEE Journal on Selected Areas in Communications, vol.
23, no. 2, pp. 201–220, February 2005.
[3] Future Directions in Cognitive Radio Network Research, NSF Workshop Report, Eds. P. Steenkiste, D. Sicker, G. Minden, and D.
Raychaudhuri, March 9-10, 2009. [Online] Available at: http://www.cs.cmu.edu/∼prs/NSF CRN Report Final.pdf
[4] J. Mitola III and G. Q. Maguire Jr., “Cognitive radio: Making software radios more personal,” IEEE Personal Communications, vol.
6, no. 4, pp. 13–18, August 1999.
[5] J. Mitola III, Cognitive radio: An Integrated Agent Architecture for Software Defined Radio, Doctor of Technology Thesis, Royal Inst.
Technol. (KTH), Stockholm, Sweden, 2000.
[6] Federal Communications Commission, “Spectrum policy task force report.” ET Docket No. 02-135, 2002.
[7] Federal Communications Commission, “In the Matter of Unlicensed Operation in the TV Broadcast Bands,” FCC 08-260, November
2008.
[8] Federal Communications Commission, “Unlicensed operation in the TV broadcast bands and additional spectrum for unlicensed
devices below 900 MHz and in the 3 GHz band,” FCC 10-174, ET Docket No. 04-186 and 02-380, September 2010.
[9] I. F. Akyildiz, W.-Y. Lee, M. C. Vuran, and S. Mohanty, “NeXt generation/dynamic spectrum access/cognitive radio wireless networks:
A survey,” Computer Networks (Elsevier), vol. 50, no. 13, 15 September 2006, pp. 2127–2159.
[10] M. Buddhikot, “Understanding dynamic spectrum access: Models, taxonomy and challenges,” in Proc. of IEEE DySPAN 2007, Dublin,
Ireland, April 17-21, 2007.
[11] Y.-C. Liang, K.-C. Chen, G. Y. Li, and P. Mahonen, “Cognitive radio networking and communications: An overview,” IEEE
Transactions on Vehicular Technology, vol. 60, no. 7, pp. 3386–3407, September 2011.
[12] B. Wang and K. J. R. Liu, “Advances in cognitive radio networks: A survey,” IEEE Journal of Selected Topics in Signal Processing,
vol. 5, no. 1, pp. 5–23, February 2011.
[13] Q. Zhao and B. M. Sadler, “A survey of dynamic spectrum access,” IEEE Signal Processing Magazine, vol. 24, no. 3, pp. 79–89,
May 2007.
[14] A. Goldsmith, S. A. Jafar, I. Maric, and S. Srinivasa, “Breaking spectrum gridlock with cognitive radios: An information theoretic
perspective,” Proceedings of the IEEE, vol. 97, no. 5, pp. 894–914, May 2009
[15] E. Hossain, D. Niyato, and Z. Han, Dynamic Spectrum Access and Management in Cognitive Radio Networks. Cambridge University
Press, 2009.
[16] Cognitive Radio Communications and Networks: Principles and Practice. (Eds. Alexander M. Wyglinski, Maziar Nekovee, and Y.
Thomas Hou), Elsevier, December 2009.
[17] Cognitive Wireless Communication Networks. (Eds. E. Hossain and V. K. Bhargava), ISBN: 978-0-387-68830-5, Springer, 2007.
[18] Cognitive Radio Networks. (Eds. Yang Xiao and Fei Hu), 978-1-4200-6420-9, CRC Press, 2009.
[19] J. Mitola III, Cognitive Radio Architecture: The Engineering Foundations of Radio XML. ISBN: 978-0471742449, Wiley Press, 2006.
[20] B. A. Fette, Cognitive Radio Technology (2nd Edition). Academic Press, ISBN: 978-0123745354, 2009.
[21] L. E. Doyle, Essentials of Cognitive Radio. Cambridge University Press, 2009.
[22] E. Biglieri, A. J. Goldsmith, L. J. Greenstein, N. Mandayam, and H. V. Poor, Principles of Cognitive Radio, Cambridge University
Press, December 2012.
[23] K.-C. Chen and R. Prasad, Cognitive Radio Networks. John Wiley and Sons, 2009.
Page 53 of 66
http://mc.manuscriptcentral.com/wcm
Wireless Communications and Mobile Computing
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
For Peer Review
52
[24] IEEE Journal on Selected Areas in Communications, Special Issue on “Cognitive Radio: Theory and Application”, vol. 26, no. 1,
January 2008.
[25] Proceedings of the IEEE, Special Issue on “Cognitive Radio - I”, vol. 97, no. 4, April 2009.
[26] Proceedings of the IEEE, Special Issue on “Cognitive Radio - II”, vol. 97, no. 5, May 2009.
[27] IEEE Journal on Selected Areas in Communications, Special Issue on “Advances in Cognitive Radio Networking and Communications
- I”, vol. 29, no. 2, February 2011.
[28] IEEE Journal on Selected Areas in Communications, Special Issue on “Advances in Cognitive Radio Networking and Communications
- II”, vol. 29, no. 4, April 2011.
[29] IEEE Transactions on Vehicular Technology, Special issue on the “Achievements and the Road Ahead: The First Decade of Cognitive
Radio”, vol. 59, no. 4, May 2010.
[30] IEEE Journal of Selected Topics in Signal Processing, Special Issue on “Signal Processing in Cooperative Cognitive Radio Systems”,
vol. 5, no. 1, January 2011.
[31] IEEE Wireless Communications Magazine, Special Issue on “Cognitive Radio Networks: A Practical Perspective”, August 2012.
[32] IEEE Vehicular Technology Magazine, Special Issue on “Applications of Cognitive Radio Networks”, June 2012.
[33] EURASIP Journal on Wireless Communications and Networking, Special Issue on “Ten Years of Cognitive Radio: State of the Art
and Perspectives”, November 2011.
[34] EURASIP Journal on Advances in Signal Processing, Special Issue on “Dynamic Spectrum Access for Wireless Networking”, August
2009.
[35] EURASIP Journal on Advances in Signal Processing, Special Issue on “Advanced Signal Processing for Cognitive Radio Networks”,
2010.
[36] Elsevier Computer Communications, Special Issue on “Cognitive Radio and Dynamic Spectrum Sharing Systems”, vol. 32, no. 18,
December 2009.
[37] R. Deng, J. Chen, X. Cao, Y. Zhang, S. Maharjan, and S. Gjessing, “Sensing-Performance Tradeoff in Cognitive Radio Enabled Smart
Grid,” IEEE Transactions on Smart Grid, vol. 4, no. 1, pp. 302-310, March 2013.
[38] J. Huang, H. Wang, Y. Qian, and C. Wang, “Priority-Based Traffic Scheduling and Utility Optimization for Cognitive Radio
Communication Infrastructure-Based Smart Grid,” IEEE Transactions on Smart Grid, vol. 4, no. 1, pp. 78-86, March 2013.
[39] Y. Zhang, R. Yu, M. Nekovee, Y. Liu, S. Xie, and S. Gjessing, “Cognitive machine-to-machine communications: visions and potentials
for the smart grid,” IEEE Network, vol. 26, no. 3, pp. 6-13, May-June 2012.
[40] S.-H. Wu, H.-L. Chao, C.-H. Ko, S.-R. Mo, C.-T. Jiang, T.-L. Li, C.-C. Cheng, and C.-F. Liang, “A cloud model and concept prototype
for cognitive radio networks,” IEEE Wireless Communications, vol. 19, no. 4, pp. 49-58, August 2012.
[41] A. Ghasemi and E. S. Sousa, “Spectrum sensing in cognitive radio networks: Requirements, challenges and design trade-offs,” IEEE
Communications Magazine, vol. 46, no. 4, pp. 32–39, April 2008.
[42] A. Ghasemi and E. S. Sousa, “Interference aggregation in spectrum-sensing cognitive wireless networks,” IEEE Journal of Selected
Topics in Signal Processing, vol. 2, no. 1, pp. 41–56, February 2008.
[43] T. Yucek and H. Arslan, “A survey of spectrum sensing algorithms for cognitive radio applications,” IEEE Communications Surveys
and Tutorials, vol. 11, no. 1, pp.116–130, 2009.
[44] Y. Zeng, Y.-C. Liang, A. T. Hoang, and R. Zhang, “A review on spectrum sensing for cognitive radio: Challenges and solutions,”
EURASIP Journal on Advances in Signal Processing, volume 2010, Article ID 381465, 2010.
[45] P. D. Sutton, K. E. Nolan, and L. E. Doyle, “Cyclostationary signatures in practical cognitive radio applications,” IEEE Journal on
Selected Areas in Communications, vol. 26, no. 1, pp. 13–24, January 2008.
Page 54 of 66
http://mc.manuscriptcentral.com/wcm
Wireless Communications and Mobile Computing
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
For Peer Review
53
[46] A. F. Molisch, M. Shafi, and L. J. Greenstein, “Propagation issues for cognitive radio,” Proceedings of the IEEE, Special Issue on
“Cognitive Radio”, 97, 787–804 (2009).
[47] A. Ghasemi and E. S. Sousa, “Fundamental limits of spectrum-sharing in fading environments,” IEEE Transactions on Wireless
Communications, vol. 6, no. 2, pp. 649–658, February 2007.
[48] Y. Zeng and Y.-C. Liang, “Eigenvalue based spectrum sensing algorithms for cognitive radio,” IEEE Transactions Communications,
vol. 57, no. 6, pp. 1784–1793, 2009.
[49] A. Mariani, A. Giorgetti, and M. Chiani, “Effects of noise power estimation on energy detection for cognitive radio applications,”
IEEE Transactions on Communications, vol. 59, no. 12, pp. 3410-3420, December 2011.
[50] R. Tandra and A. Sahai, “SNR walls for signal detection,” IEEE Journal of Selected Topics in Signal Processing, vol. 2, no. 1, pp.
4–17, February 2008.
[51] S. Haykin, D. Thomson, and J. Reed, “Spectrum sensing for cognitive radio,” Proceedings of the IEEE, Special Issue on “Cognitive
Radio”, vol. 97, no. 5, pp. 849–877, May 2009.
[52] Y.-C. Liang, Y. Zeng, E. C. Y. Peh, and A. T. Hoang, “Sensing-throughput tradeoff for cognitive radio networks,” IEEE Transactions
on Wireless Communications, vol. 7, no. 4, pp. 1326–1337, April 2008.
[53] K. B. Letaief and W. Zhang, “Cooperative communications for cognitive radio networks,” Proceedings of the IEEE, vol. 97, no. 5,
pp. 878–893, May 2009.
[54] G. Ganesan and Y. Li, “Cooperative spectrum sensing in cognitive radio, part I: Two user networks,” IEEE Transactions on Wireless
Communications, vol. 6, no. 6, pp. 2204–2213, June 2007.
[55] G. Ganesan and Y. Li, “Cooperative spectrum sensing in cognitive radio, part II: Multiuser networks,” IEEE Transactions on Wireless
Communications, vol. 6, no. 6, pp. 2214–2222, June 2007.
[56] J. Unnikrishnan and V. V. Veeravalli, “Cooperative sensing for primary detection in cognitive radio,” IEEE Journal of Selected Topics
in Signal Processing, vol. 2, no. 1, pp. 18–27, February 2008.
[57] R. Zhang and Y.-C. Liang, “Exploiting multi-antennas for opportunistic spectrum sharing in cognitive radio networks,” IEEE Journal
of Selected Topics in Signal Processing, vol. 2, no. 1, pp. 88–102, February 2008.
[58] Z. Quan, S. Cui, and A. H. Sayed, “Optimal linear cooperation for spectrum sensing in cognitive radio networks,” IEEE Journal of
Selected Topics in Signal Processing, vol. 2, no. 1, pp. 28–40, February 2008.
[59] J. Meng, W. Yin, H. Li, E. Hossain, and Z. Han, “Collaborative spectrum sensing from sparse observations in cognitive radio
networks,” IEEE Journal on Selected Topics on Communications, Special Issue on “Advances in Cognitive Radio Networking and
Communications”, vol. 29, no. 2, pp. 327–337, February 2011.
[60] K. W. Choi, E. Hossain, and D. I. Kim, “Cooperative spectrum sensing under random geometric primary user network model,” IEEE
Transactions on Wireless Communications, vol. 10, no. 6, pp. 1932–1944, June 2011.
[61] L. Lu, X. Zhou, U. Onunkwo, and G. Y. Li, “Ten years of research in spectrum sensing and sharing in cognitive radio,” EURASIP
Journal on Wireless Communications and Networking, 2012:28, January 2012.
[62] G. L. Stuber, S. M. Almalfouh, and D. Sale, “Interference analysis of TV-band white space,” Proceedings of the IEEE, vol. 97, no.
4, pp. 741–754, April 2009.
[63] A. Rabbachin, Tony Q. S. Quek, H. Shin, and M. Z. Win, “Cognitive network interference,” IEEE Journal on Selected Areas in
Communications, vol. 29, no. 2, February 2011.
[64] D. A. Roberson, C. S. Hood, J. L. LoCicero, and J. T. MacDonald, “Spectral occupancy and interference studies in support of cognitive
radio technology deployment,” in Proc. of First IEEE Workshop on Networking Technologies for Software Defined Radio Networks,
pp. 26–35, September 2006.
Page 55 of 66
http://mc.manuscriptcentral.com/wcm
Wireless Communications and Mobile Computing
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
For Peer Review
54
[65] D. Datla, A. M. Wyglinski, and G. J. Minden, “A spectrum surveying framework for dynamic spectrum access networks,” IEEE
Transactions on Vehicular Technology, vol. 58, no. 8, pp. 4158–4158, April 2009.
[66] C. Ghosh, S. Pagadarai, D. Agrawal, and A. M. Wyglinski, “A framework for statistical wireless spectrum occupancy modeling,”
IEEE Transactions on Wireless Communications, vol. 9, no. 1, pp. 38–44, January 2010.
[67] M. Wellens and P. Mahonen, “Lessons learned from an extensive spectrum occupancy measurement campaign and a stochastic duty
cycle model,” Mobile Networks and Applications (Springer), 2009.
[68] B. Canberk, I. F. Akyildiz, and S. Oktug, “Primary user activity modeling using first-difference filter clustering and correlation in
cognitive radio networks,” IEEE/ACM Transactions on Networking, vol. 19, no. 1, pp. 170–183, February 2011.
[69] T. Jiang, Z.-G. Cui, J.-W. Hu, and Y.-L. Hou, “A design of cognitive UWB pulse based on APSWF,” in Proc. of Asia-Pacific Symposium
on Electromagnetic Compatibility (APEMC 2008), pp. 706–709, 19-23 May 2008.
[70] W. Zhang, H. Shen, Z. Bai, and K. S. Kwak, “A novel modulation waveform on ultra-wideband based cognitive radio systems,” in
Proc. of 2nd International Conference on Cognitive Radio Oriented Wireless Networks and Communications (CrownCom 2007), pp.
306–311, 1-3 August 2007.
[71] F. Wei, P. Xia, Z. Yang, and F. Tian, “Decentralized waveform design for MIMO cognitive radio under interference temperature
constraint,” in Proc. of 2011 Second International Conference on Networking and Distributed Computing (ICNDC), pp.159–162,
21-24 September 2011.
[72] L.-L. Zhou, H.-B. Zhu, and N.-T. Zhang, “Iterative solution to the notched waveform design in cognitive ultra-wideband radio system,”
Progress In Electromagnetic Research, vol. 75, pp. 271–284, 2007.
[73] Z. Tian, G. Leus, and V. Lottici, “Joint dynamic resource allocation and waveform adaptation in cognitive radio networks,” in Proc.
of IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP 2008), pp. 5368–5371, March 31 2008-April
4 2008.
[74] V. Chakravarthy, X. Li, Z. Wu, M. Temple, F. Garber, R. Kannan, and A. Vasilakos, “Novel overlay/underlay cognitive radio waveforms
using SD-SMSE framework to enhance spectrum efficiency- part i: Theoretical framework and analysis in AWGN channel,” IEEE
Transactions on Communications, vol. 57, no. 12, pp. 3794–3804, December 2009.
[75] S. Pagadarai, A. Kliks, H. Bogucka, and A. M. Wyglinski, “On non-contiguous multicarrier waveforms for spectrally opportunistic
cognitive radio systems,” in Proc. of 2010 International Waveform Diversity and Design Conference (WDD), pp. 177–181, 8-13 August
2010.
[76] Z. Hu, N. Guo, and R. Qiu, “Wideband waveform design for relay cognitive network,” in Proc. of IEEE Military Communications
Conference (Milcom 2010), pp. 749–754, October 31-November 3, 2010.
[77] Z. Hu, N. Guo, and R. Qiu, “Wideband waveform optimization for multiple input single output cognitive radio with practical
considerations,” in Proc. of IEEE Military Communications Conference (Milcom 2010), pp. 1227–1232, October 31-November 3,
2010.
[78] Z. Kollar and P. Horvath, “Physical layer considerations for cognitive radio: Modulation techniques,” in Proc. of 2011 IEEE 73rd
Vehicular Technology Conference (VTC Spring), pp. 1–5, 15-18 May 2011.
[79] Y. Chen, M.-S. Alouini, and L. Tang, “Performance analysis of adaptive modulation for cognitive radios with opportunistic access,”
in Proc. of 2011 IEEE International Conference on Communications (ICC), pp. 1–5, 5-9 June 2011.
[80] M. Zamanian, A. A. Tadaion, and M. T. Sadeghi, “Modulation classification of linearly modulated signals in a cognitive radio
network using constellation shape,” in Proc. of 2011 7th International Workshop on Systems, Signal Processing and their Applications
(WOSSPA), pp. 13–16, 9-11 May 2011.
[81] M. R. Khanzadi, K. Haghighi, A. Panahi, and T. Eriksson, “A novel cognitive modulation method considering the performance of
Page 56 of 66
http://mc.manuscriptcentral.com/wcm
Wireless Communications and Mobile Computing
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
For Peer Review
55
primary user,” in Proc. of 2010 6th Conference on Wireless Advanced (WiAD), pp. 1–6, 27-29 June 2010.
[82] Z. Ye, G. Memik, and J. Grosspietsch, “Digital modulation classification using temporal waveform features for cognitive radios,”
Proc. of IEEE Int. Symposium on Personal, Indoor and Mobile Radio Communications (PIMRC 2007), pp. 1–5, 3-7 September 2007.
[83] I. Budiarjo, H. Nikookar, and L. Ligthart, “Cognitive radio modulation techniques,” IEEE Signal Processing Magazine, vol. 25, no.
6, pp. 24–34, November 2008.
[84] R. Rajbanshi, Q. Chen, A. M. Wyglinski, G. J. Minden, and J. B. Evans, “Quantitative comparison of agile modulation techniques for
cognitive radio transceivers,” in Proc. of 4th IEEE Consumer Communications and Networking Conference (CCNC 2007), pp. 1144–
1148, 11–13 January 2007.
[85] P. Kryszkiewicz, H. Bogucka, and A. M. Wyglinski, “Protection of primary users in dynamically varying radio environment: Practical
solutions and challenges,” EURASIP Journal on Wireless Communications and Networking, 2012:23.
[86] D. Qu, Z. Wang, and T. Jiang, “Extended active interference cancellation for sidelobe suppression in cognitive radio OFDM systems
with cyclic prefix,” IEEE Transactions on Vehicular Technology, vol. 59, no. 4, pp.1689–1695, May 2010.
[87] A. Ghasemi and E. S. Sousa, “Interference aggregation in spectrum-sensing cognitive wireless networks,” IEEE Journal of Selected
Topics in Signal Processing, vol. 2, no. 1, pp. 41–56, February 2008.
[88] R. Rajbanshi, Q. Chen, A. M. Wyglinski, J. B. Evans, and G. J. Minden, “Comparative study of frequency agile data transmis- sion
schemes for cognitive radio transceivers,” in Proc. of 1st Int. Wksp. on Technol. and Policy for Accessing Spectrum, Boston, MA,
USA, August 2006.
[89] A. Batra, S. Lingam, and J. Balakrishnan, “Multi-band OFDM: A cognitive radio for UWB,” in Proc. of IEEE International Symposium
on Circuits and Systems (ISCAS 2006), 21-24 May 2006.
[90] S.-M. Cheng, W. C. Ao, and K.-C. Chen, “Efficiency of a cognitive radio link with opportunistic interference mitigation,” IEEE
Transactions on Wireless Communications, vol. 10, no. 6, pp. 1715–1720, June 2011.
[91] H. Bogucka, A. M. Wyglinski, S. Pagadarai, A. Kliks, “Spectrally agile multicarrier waveforms for opportunistic wireless access,”
IEEE Communications Magazine, vol. 49, no. 6, pp. 108–115, June 2011.
[92] M. M. Ramon, T. Atwood, S. Barbin, and C. G. Christodoulou, “Signal classification with an SVM-FFT approach for feature extraction
in cognitive radio,” in Proc. of SBMO/IEEE MTT-S International Microwave & Optoelectronics Conference (IMOC 2009).
[93] M. G. Khoshkholgh, K. Navaie, and H. Yanikomeroglu, “Access strategies for spectrum sharing in fading environment: Overlay,
underlay and mixed,” IEEE Transactions on Mobile Computing, vol. 9, no. 12, pp. 1780–1793, December 2010.
[94] R. A. Tannious and A. Nosratinia, “Cognitive radio protocols based on exploiting hybrid ARQ retransmissions,” IEEE Transactions
on Wireless Communications, vol. 9, no. 9, pp. 2833–2841, September 2010.
[95] Y. Xing, R. Chandramouli, S. Mangold, and S. Sankar N, “Dynamic spectrum access in open spectrum wireless networks,” IEEE
Journal on Selected Areas in Communications, vol. 24, no. 3, pp. 626-637, March 2006.
[96] L. B. Le and E. Hossain, “Resource allocation for spectrum underlay in cognitive wireless networks,” IEEE Transactions on Wireless
Communications, vol. 7, no. 12, December 2008, pp. 5306–5315.
[97] Y.-C. Liang, Y. Zeng, E.C.Y. Peh, and A. T. Hoang, “Sensing-throughput tradeoff for cognitive radio networks,” IEEE Transactions
on Wireless Communications, vol. 7, no. 4, pp. 1326–1337, April 2008.
[98] D. I. Kim, L. B. Le, and E. Hossain, “Joint rate and power allocation for cognitive radios in dynamic spectrum access environment,”
IEEE Transactions on Wireless Communications, vol. 7, no. 12 - part 2, pp. 5517–5527, December 2008.
[99] R. Zhang and Y.-C. Liang, “Exploiting multi-antennas for opportunistic spectrum sharing in cognitive radio networks,” IEEE Journal
of Selected Topics in Signal Processing, vol. 2, no.1, pp. 88–102, February 2008.
Page 57 of 66
http://mc.manuscriptcentral.com/wcm
Wireless Communications and Mobile Computing
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
For Peer Review
56
[100] L. Zhang, Y.-C. Liang, and Y. Xin, “Joint beamforming and power allocation for multiple access channels in cognitive radio networks,”
IEEE Journal on Selected Areas in Communications, vol. 26, no.1, pp. 38–51, January 2008.
[101] R. Zhang, Y.-C. Liang and S. Cui, “Dynamic resource allocation in cognitive radio networks,” IEEE Signal Processing Magazine,
vol. 27, no. 3, pp. 102–114, May 2010.
[102] E. Hossain, L. Le, N. Devroye, and M. Vu, “Cognitive radio: From theory to practical network engineering,” invited chapter in
Advances in Wireless Communications, (Ed. V. Tarokh), Springer, 2009.
[103] S. C. Jha, M. M. Rashid, V. K. Bhargava, and C. L. Despins, “Medium access control in distributed cognitive radio networks,” IEEE
Wireless Communications, August 2011, pp. 41–51.
[104] H. Kim and K. G. Shin, “Efficient discovery of Spectrum opportunities with MAC-layer sensing in cognitive radio networks,” IEEE
Transactions on Mobile Computing, vol. 7, no. 5, pp. 533–545, May 2008.
[105] Q. Zhao, L. Tong, A. Swami, and Y. Chen, “Decentralized cognitive MAC for opportunistic spectrum access in ad hoc networks: A
POMDP framework,” IEEE Journal on Selected Areas in Communications, vol. 25, no. 3, pp. 589–600, April 2007.
[106] J. Jia, Q. Zhang, and X. Shen, “HC-MAC: A hardware-constrained cognitive MAC for efficient spectrum management,” IEEE Journal
on Selected Areas in Communications, vol. 26, no. 1, pp. 106–117, January 2008.
[107] S. Wang and J. Zhang, “Opportunistic Spectrum Scheduling by Jointly Exploiting Channel Correlation and PU Traffic Memory,” IEEE
Journal on Selected Areas in Communications, vol. 31, no. 3, pp. 394-405, March 2013.
[108] Y. Chen, Q. Zhao, and A. Swami, “Joint design and separation principle for opportunistic spectrum access in the presence of sensing
errors,” IEEE Transactions on Information Theory, vol. 54, no. 5, pp. 2053–2071, May 2008.
[109] Y. Xing, R. Chandramouli, S. Mangold, and S. Sankar N, “Dynamic spectrum access in open spectrum wireless networks,” IEEE
Journal on Selected Areas in Communications, vol. 24, no. 3, pp. 626–637, March 2006.
[110] Y. Xing, C. N. Mathur, M. A. Haleem, R. Chandramouli, and K. P. Subbalakshmi, “Dynamic spectrum access with QoS and interference
temperature constraints,” IEEE Transactions on Mobile Computing, vol. 6, no. 4, pp. 423–433, April 2007.
[111] H. Su and X. Zhang, “Cross-layer based opportunistic MAC protocols for QoS provisioning over cognitive radio wireless networks,”
IEEE Journal on Selected Areas in Communications, vol. 26, no. 1, pp. 118–129, January 2008.
[112] H. Jiang, L. Lai, R. Fan, and H. V. Poor, “Optimal selection of channel sensing order in cognitive radio,” IEEE Transactions on
Wireless Communications, vol. 8, no. 1, pp. 297–303, January 2009.
[113] H. Su and X. Zhang, “Cross-layer based opportunistic MAC protocols for QoS provisionings over cognitive radio wireless networks,”
IEEE Journal on Selected Areas in Communications, vol. 26, no. 1, pp. 118-129, Jan. 2008.
[114] B. Jalaeian and M. Motani, “Location Aware CR-MAC: A multi-channel cross layered PHY-MAC protocol for cognitive radio ad hoc
networks,” in Proceedings of IEEE Malaysia International Conference on Communications (MICC), pp. 348-353, 15-17 Dec. 2009.
[115] W. S. Jeon, D. H. Lee, and D. G. Jeong, “Collaborative Sensing Management for Cognitive Radio Networks with Reporting Overhead,”
IEEE Transactions on Wireless Communications, vol. 12, no. 2, pp. 595-605, February 2013.
[116] X. Kang, Y.-C. Liang, A. Nallanathan, H. K. Garg, and R. Zhang, “Optimal power allocation for fading channels in cognitive radio
networks: Ergodic capacity and outage capacity,” IEEE Transactions on Wireless Communications, vol. 8, no. 2, pp. 940–950, February
2009.
[117] G. Bansal, M. J. Hossain, and V. K. Bhargava, “Optimal and suboptimal power allocation schemes for OFDM-based cognitive radio
systems,” IEEE Trans. Wireless Commun., 7(11): 4710–4718, November 2008.
[118] G. Bansal, M. J. Hossain, and V. K. Bhargava, “Adaptive power loading for OFDM-based cognitive radio systems with statistical
interference constraint,” IEEE Trans. Wireless Commun., vol. 10, no. 9, pp. 2786–2791, September 2011.
Page 58 of 66
http://mc.manuscriptcentral.com/wcm
Wireless Communications and Mobile Computing
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
For Peer Review
57
[119] D. Bharadia, G. Bansal, P. Kaligineedi, and V. K. Bhargava, “Relay and power allocation schemes for OFDM-based cognitive radio
systems,” IEEE Trans. Wireless Commun., vol. 10, no. 9, pp. 2812–2817, September 2011.
[120] Z. Hasan, G. Bansal, E. Hossain, and V. K. Bhargava, “Energy-efficient power allocation in OFDM-based cognitive radio systems: A
risk-return model,” IEEE Trans. Wireless Commun., 8(12): 6078–6088, December 2009.
[121] P. Phunchongharn and E. Hossain, “Distributed robust scheduling and power control for cognitive spatial-reuse TDMA networks,”
IEEE Journal on Selected Areas in Communications, vol. 30, no. 10, pp. 1934-1946, November 2012.
[122] M. M. Rashid, M. J. Hossain, E. Hossain, and V. K. Bhargava, “Opportunistic spectrum scheduling for multiuser cognitive radio: A
queueing analysis,” IEEE Trans. Wireless Commun., 8(10): 5259–5269, October 2009.
[123] I. F. Akyildiz, Won-Yeol Lee, M. C. Vuran, and S. Mohanty, “A survey on spectrum management in cognitive radio networks,” IEEE
Communications Magazine, vol. 46, no. 4, pp. 40–48, April 2008.
[124] L.-C. Wang, C.-W. Wang, and K.-T. Feng, “A queueing-theoretical framework for QoS-enhanced spectrum management in cognitive
radio networks,” IEEE Wireless Communications, vol. 18, no. 6, pp. 18–26, December 2011.
[125] R. Zhou, X. Li, V. Chakravarthy, and Z. Wu, “Spectrum mobility demonstration of SMSE based overlay cognitive radio via software
defined radio,” in Proc. of IEEE Symposium on New Frontiers in Dynamic Spectrum Access Networks (DySPAN), pp. 668–669, May
2011.
[126] R. W. Thomas, D. H. Friend, L. A. Dasilva, and A. B. Mackenzie, “Cognitive networks: Adaptation and learning to achieve end-to-end
performance objectives,” IEEE Communications Magazine, vol. 44, no. 12, pp. 51–57, December 2006.
[127] N. Nie and C. Comaniciu, “Adaptive channel allocation spectrum etiquette for cognitive radio networks,” in Proc. of First IEEE
International Symposium on Dynamic Spectrum Access Networks (DySPAN05), pp. 269–278, November 2005.
[128] J.-S. Pang and G. Scutari, “Joint Sensing and Power Allocation in Nonconvex Cognitive Radio Games: Quasi-Nash Equilibria,” IEEE
Transactions on Signal Processing, vol. 61, no. 9, pp. 2366-2382, May 2013.
[129] A. He, K. K. Bae, T. Newman, J. Gaeddert, K. Kim, R. Menon, L. Morales-Tirado, J. Neel, Y. Zhao, J. Reed, and W. Tranter, “A
survey of artificial intelligence for cognitive radios,” IEEE Transactions on Vehicular Technology, vol. 59, no. 4, pp. 1578–1592, May
2010.
[130] Y. Xing and R. Chandramouli, “Human behavior inspired cognitive radio network design,” IEEE Communications Magazine, vol. 46,
no. 12, pp. 122–127, December 2008.
[131] C. Clancy, J. Hecker, E. Stuntebeck, and T. OShea, “Applications of machine learning to cognitive radio networks,” IEEE Wireless
Communications, vol. 14, no. 4, pp. 47–52.
[132] A. G.-Serrano and L. Giupponi, “Distributed Q-learning for aggregated interference control in cognitive radio networks,” IEEE
Transactions on Vehicular Technology, vol. 59, no. 4, pp. 1823–1834, May 2010.
[133] Z. Han, R. Zheng, and H. Poor, “Repeated auctions with bayesian nonparametric learning for spectrum access in cognitive radio
networks,” IEEE Transactions on Wireless Communications, vol. 10, no. 3, pp. 890–900, Mar. 2011.
[134] T. Clancy, A. Khawar, and T. Newman, “Robust signal classification using unsupervised learning,” IEEE Transactions on Wireless
Communications, vol. 10, no. 4, pp. 1289–1299, April 2011.
[135] M. Maskery, V. Krishnamurthy, and Q. Zhao, “Decentralized dynamic spectrum access for cognitive radios: Cooperative design of a
non-cooperative game,” IEEE Transactions on Communications, vol. 57, no. 2, pp. 459–469, February 2009.
[136] M. van der Schaar and F. Fu, “Spectrum access games and strategic learning in cognitive radio networks for delay-critical applications,”
Proceedings of the IEEE, vol. 97, no. 4, pp. 720–740, April 2009.
[137] N. Baldo, B. Tamma, B. Manojt, R. Rao, and M. Zorzi, “A neural network based cognitive controller for dynamic channel selection,”
in Proc. of IEEE International Conference on Communications (ICC09), June 2009, pp. 1–5.
Page 59 of 66
http://mc.manuscriptcentral.com/wcm
Wireless Communications and Mobile Computing
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
For Peer Review
58
[138] V. Tumuluru, P. Wang, and D. Niyato, “A neural network based spectrum prediction scheme for cognitive radio,” in Proc. of IEEE
International Conference on Communications (ICC’10), May 2010, pp. 1–5.
[139] A. Anandkumar, N. Michael, A. K. Tang, and A. Swami, “Distributed algorithms for learning and cognitive medium access with
logarithmic regret,” IEEE JSAC on Advances in Cognitive Radio Networking and Communications, vol. 29, no. 4, pp. 781–745, April
2011.
[140] K. W. Choi and E. Hossain, “Opportunistic access to spectrum holes between packet bursts: A learning-based approach,” IEEE
Transactions on Wireless Communications, vol. 10, no. 8, pp. 2497–2509, August 2011.
[141] I. F. Akyildiz, W.-Y. Lee, and K. R. Chowdhury, “CRAHNs: Cognitive radio ad hoc networks,” Ad Hoc Networks (Elsevier), vol. 7,
no. 5, July 2009, pp. 810–836.
[142] Z. Yang, G. Cheng, W. Liu, W. Yuan, W. Cheng, “Local coordination based routing and spectrum assignment in multi-hop cognitive
radio networks,” ACM MONET, vol. 13, pp. 67–81, April 2008.
[143] Y. T. Hou, Y. Shi, and H. D. Sherali, “Spectrum sharing for multi-hop networking with cognitive radios,” IEEE Journal on Selected
Areas in Communications, vol. 26, no. 1, pp. 146–155, January 2008.
[144] R. Urgaonkar and M. J. Neely, “Opportunistic scheduling with reliability guarantees in cognitive radio networks,” IEEE Transactions
on Mobile Computing, vol. 8, no. 6, pp. 766–777, June 2009.
[145] D. Xue and E. Ekici, “Cross-Layer Scheduling for Cooperative Multi-Hop Cognitive Radio Networks,” IEEE Journal on Selected
Areas in Communications, vol. 31, no. 3, pp. 534-543, March 2013.
[146] S. H. Song, M. O. Hasna, and K. B. Letaief, “Prior Zero Forcing for Cognitive Relaying,” IEEE Transactions on Wireless
Communications, vol. 12, no. 2, pp. 938-947, February 2013.
[147] J.-C. Liang and J.-C. Chen, “Resource Allocation in Cognitive Radio Relay Networks,” IEEE Journal on Selected Areas in
Communications, vol. 31, no. 3, pp. 476-488, March 2013.
[148] Y. Wu and D. H. K. Tsang, “Dynamic rate allocation, routing and spectrum sharing for multi-hop cognitive radio networks,” IEEE
International Conference on Communications (ICC’09) Workshops, pp. 1–6, June 2009.
[149] T. C. Clancy and N. Goergen, “Security in cognitive radio networks: Threats and mitigation,” in Proc. of 3rd International Conference
on Cognitive Radio Oriented Wireless Networks and Communications (CrownCom 2008), pp. 1–8, May 2008.
[150] R. Chen, J.-M. Park, and J. H. Reed, “Defense against primary user emulation attacks in cognitive radio networks,” IEEE Journal on
Selected Areas in Communications, vol. 26, no. 1, pp. 25–37, January 2008.
[151] R. Chen, J.-M. Park, Y. T. Hou, and J. H. Reed, “Toward secure distributed spectrum sensing in cognitive radio networks,” IEEE
Communications Magazine, vol. 46, no. 4, pp. 50–55, April 2008.
[152] S. Zheng, P.-Y. Kam, Y.-C. Liang, and Y. Zeng, “Spectrum sensing for digital primary signals in cognitive radio: A Bayesian approach
for maximizing spectrum utilization,” IEEE Transactions on Wireless Communications, vol. 12, no. 4, pp. 1774-1782, April 2013.
[153] H. Li and Z. Han, “Dogfight in spectrum: Combating primary user emulation attacks in cognitive radio systems, part I: Known channel
statistics,” IEEE Transactions on Wireless Communications, vol. 9, no. 11, pp. 3566–3577, November 2010
[154] H. Li and Z. Han, “Dogfight in spectrum: Combating primary user emulation attacks in cognitive radio systems, part II: Unknown
channel statistics,” IEEE Transactions on Wireless Communications, vol. 10, no. 1, pp. 274–283, January 2011.
[155] H. Li and Z. Han, “Catch me if you can: An abnormality detection approach for collaborative spectrum sensing in cognitive radio
networks,” IEEE Transactions on Wireless Communications, vol. 9, no.11, pp. 3554–3565, November 2010.
[156] D. Niyato and E. Hossain, “Spectrum trading in cognitive radio networks: A market-equilibrium-based approach,” IEEE Wireless
Communications Magazine, vol. 15, no. 6, pp. 71–80, December 2008.
Page 60 of 66
http://mc.manuscriptcentral.com/wcm
Wireless Communications and Mobile Computing
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
For Peer Review
59
[157] J. Huang, R. Berry, and M. Honig, “Auction-based spectrum sharing,” Mobile Networks and Applications, vol. 11, no. 3, pp. 405–418,
June 2006.
[158] J. Sun, E. Modiano, and L. Zheng, “Wireless channel allocation using an auction algorithm,”, IEEE Journal on Selected Areas in
Communications, vol. 24, no. 5, pp. 1085–1096, May 2006.
[159] Y. Chen, Y. Wu, B. Wang, and K. J. R. Liu, “Spectrum auction games for multimedia streaming over cognitive radio networks,” IEEE
Transactions on Communications, vol. 58, no. 8, pp. 2381–2390, August 2010.
[160] G. S. Kasbekar and S. Sarkar, “Spectrum auction framework for access allocation in cognitive radio networks,” IEEE/ACM Transactions
on Networking, vol. 18, no. 6, pp. 1841–1854, December 2010.
[161] S. Sodagari, A. Attar, and S. G. Bilen, “On a truthful mechanism for expiring spectrum sharing in cognitive radio networks,” IEEE
Journal on Selected Areas in Communications, vol. 29, no. 4, pp. 856–865, April 2011.
[162] Z. Han, R. Zheng, and H. V. Poor, “Repeated auctions with Bayesian nonparametric learning for spectrum access in cognitive radio
networks,” IEEE Transactions on Wireless Communications, vol. 10, no. 3, pp. 890–900, March 2011.
[163] F. Fu and M. van der Schaar, “Learning to compete for resources in wireless stochastic games,” IEEE Transactions on Vehicular
Technology, vol. 58, no. 4, pp. 1904–1919, May 2009.
[164] D. Niyato and E. Hossain, “Competitive pricing for spectrum sharing in cognitive radio networks: Dynamic game, inefficiency of
Nash equilibrium, and collusion,” IEEE Journal on Selected Areas in Communications, vol. 26, no. 1, pp. 192–202, January 2008
[165] F. Wang, M. Krunz, and S. Cui, “Price-based spectrum management in cognitive radio networks,” IEEE Journal on Selected Topics
in Signal Processing, vol. 2, no. 1, pp. 74–87, February 2008.
[166] Z. Ji and K. J. R. Liu, “Dynamic spectrum sharing: A game theoretical overview,” IEEE Communications Magazine, vol. 45, no. 5,
May 2007.
[167] D. Niyato and E. Hossain, “Competitive spectrum sharing in cognitive radio networks: A dynamic game approach,” IEEE Transactions
on Wireless Communications, vol. 7, no. 7, July 2008, pp. 2651–2660.
[168] D. Niyato, E. Hossain, and Z. Han, “Dynamics of multiple-seller and multiple-buyer spectrum trading in cognitive radio networks: A
game theoretic modeling approach,” IEEE Transactions on Mobile Computing, vol. 8, no. 8, August 2009, pp. 1009–1022.
[169] Z. Ji and K. J. R. Liu, “Multi-stage pricing game for collusion-resistant dynamic spectrum allocation,” IEEE Journal on Selected
Areas in Communications, vol. 26, no. 1, pp. 182–191, January 2008.
[170] L. Yang, H. Kim, J. Zhang, M. Chiang, and C. W. Tan, “Pricing-based decentralized spectrum access control in cognitive radio
networks,” IEEE/ACM Transactions on Networking, vol. 21, no. 2, pp. 522-535, April 2013.
[171] S. M. M. Toroujeni, S. M. S. Sadough, and S. A. G. Ghorashi, “Spectrum Leasing for OFDM-Based Cognitive Radio Networks,”
IEEE Transactions on Vehicular Technology, vol. 62, no. 5, pp. 2131-2139, Jun 2013.
[172] A. Pereira, H. Fontes, and A. Gameiro, “Simulation platform for opportunistic radio systems,” in Proc. of IEEE International Symposium
on Personal, Indoor and Mobile Radio Communications (PIMRC), September 2006.
[173] M. Zhan, P. Ren, and M. Gong, “An open software simulation platform for cognitive radio,” in Proc. of International Conference on
Wireless Communications Networking and Mobile Computing (WiCOM), September 2010.
[174] A. Sibille, “Efficient generation of spatially and frequency correlated random values for cognitive radio network simulators,” IEEE
Transactions on Vehicular Technology, vol. 59, no. 3, pp. 1121–1128, March 2010.
[175] A. A. Abidi, “The path to the software-defined radio receiver,” IEEE Journal of Solid-State Circuits, vol. 42, no. 5, pp. 954–966, May
2007.
[176] J. Mitola, “The software radio architecture,” IEEE Communications Magazine, vol. 33, no. 5, pp. 26–38, May 1995.
Page 61 of 66
http://mc.manuscriptcentral.com/wcm
Wireless Communications and Mobile Computing
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
For Peer Review
60
[177] J. Mitola III, “Software radio architecture: A mathematical perspective,” IEEE Journal on Selected Areas in Communications, vol.
17, no. 4, pp. 514–538, April 1999.
[178] S. M. Mishra, D. Cabric, D. Chang, D. Willkomm, B. van Schewick, S. Wolisz, and B. W. Brodersen, “A real time cognitive radio
testbed for physical and link layer experiments,” in Proc. of First IEEE International Symposium on New Frontiers in Dynamic
Spectrum Access Networks (DySPAN05), pp. 562–567, 8-11 November 2005.
[179] D. Cabric, I. D. O’Donnell, M. S.-W. Chen, and R. W. Brodersen, “Spectrum sharing radios,” IEEE Circuits and Systems Magazine,
vol. 6, no. 2, pp. 30–45, 2006.
[180] R. Bagheri, A. Mirzaei, S. Chehrazi, M. E. Heidari, M. Lee, M. Mikhemar, W. Tang, and A. A. Abidi, “An 800 MHz - 6 GHz
software-defined wireless receiver in 90-nm CMOS,” IEEE Journal of Solid-State Circuits, vol. 41, no. 12, pp. 2860–2876, December
2006.
[181] http://www.cs.cmu.edu/∼emulator
[182] http://www.orbit-lab.org
[183] http://prisms.cs.umass.edu/dome/umassdieselnet
[184] http://www.citysense.net/
[185] http://www.winlab.rutgers.edu/docs/focus/WiNC2R.html
[186] G. J. Minden, J. B. Evans, L. Searl, D. DePardo, V. R. Petty, R. Rajbanshi, J. Guffey, Q. Chen, T. Newman, F. Weidling, D. Datla,
B. Barker, M. Peck, B. Cordill, A. M. Wyglinski, and A. Agah, “KUAR: A flexible software-defined radio development platform,” in
Proc. of Second IEEE Symposium on New Frontiers in Dynamic Spectrum Access Networks, Dublin, Ireland, April 2007.
[187] P. Marshall, “Extending the reach of cognitive radio,” Proceedings of the IEEE, vol. 97, no. 4, April 2009, pp. 612–625.
[188] GNU Radio Project, http://www.gnu.org/software/gnuradio
[189] WARP: Wireless Open Access Research Platform, 2009. http://warp.rice.edu
[190] J. N. Laneman, D. N. C. Tse, and G. W. Wornell, “Cooperative diversity in wireless networks: Efficient protocols and outage behavior,”
IEEE Trans. Info. Theory, vol. 50, no. 12, pp. 3062–3080, December 2004.
[191] Q. Zhang, J. Jia, and J. Zhang, “Cooperative relay to improve diversity in cognitive radio networks,” IEEE Communications Magazine,
vol. 47, no. 2, pp. 111–117, February 2009.
[192] H. Min, S. Lee, K. Kwak, and D. Hong, “Effect of multiple antennas at the source on outage probability for amplify-and-forward
relaying systems,” IEEE Trans. on Wireless Commun., vol. 8, no. 2, pp. 633–637, February 2009.
[193] K. B. Letaief and W. Zhang, “Cooperative communications for cognitive radio networks,” Proceedings of the IEEE, vol. 97, no. 5,
pp. 878 –893, May 2009.
[194] J. Jia, J. Zhang, and Q. Zhang, “Cooperative relay for cognitive radio networks,” IEEE INFOCOM 2009, pp. 2304–2312, April 2009.
[195] X. Gong et al., “A cooperative relay scheme for secondary communication in cognitive radio networks,” IEEE Global Telecommuni-
cations Conference (GLOBECOM 2008), pp. 1–6. December 2008.
[196] G. Zhao et al., “Spatial spectrum holes for cognitive radio with relay-assisted directional transmission,” IEEE Transactions on Wireless
Communications, vol. 8, no. 10, pp. 5270 –5279, October, 2009.
[197] J. Jia, J. Zhang, and Q. Zhang, “Relay-assisted routing in cognitive radio networks,” IEEE International Conference on Communications
(ICC’09), pp. 1–5, June 2009.
[198] Y. Zou et al., “A cognitive transmission scheme with the best relay selection in cognitive radio networks,” IEEE Global
Telecommunications Conference (GLOBECOM 2010), pp. 1–5, December 2010.
[199] L. Giupponi and C. Ibars, “Bayesian potential games to model cooperation for cognitive radios with incomplete information,” in IEEE
International Conference on Communications (ICC’09), pp. 1–6, June 2009.
Page 62 of 66
http://mc.manuscriptcentral.com/wcm
Wireless Communications and Mobile Computing
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
For Peer Review
61
[200] Y. Zou et al., “An adaptive cooperation diversity scheme with best-relay selection in cognitive radio networks,” IEEE Transactions
on Signal Processing, vol. 58, no. 10, pp. 5438–5445, October 2010.
[201] Y. Li, R. Zhang, C. Wang, A. Daneshmand, and X. You, “A distributed relay selection algorithm for cognitive radio ad-hoc networks,”
2010 International Conference on Network and Service Management (CNSM), pp. 439–442, October 2010.
[202] J. Zhu, J. Huang, and W. Zhang, “Optimal one-dimensional relay placement in cognitive radio networks,” 2010 IEEE International
Conference on Wireless Communications and Signal Processing (WCSP), pp. 1–6, October 2010.
[203] A. Rao, H. Ma, M.-S. Alouini, and Y. Chen, “Impact of Primary User Traffic on Adaptive Transmission for Cognitive Radio with
Partial Relay Selection,” IEEE Transactions on Wireless Communications, vol. 12, no. 3, pp. 1162-1172, March 2013.
[204] V. N. Q. Bao, T. Q. Duong, D. Benevides da Costa, G. C. Alexandropoulos, and A. Nallanathan, “Cognitive Amplify-and-Forward
Relaying with Best Relay Selection in Non-Identical Rayleigh Fading,” IEEE Communications Letters, vol. 17, no. 3, pp. 475-478,
March 2013.
[205] Y. Li and A. Nosratinia, “Spectrum Sharing with Distributed Relay Selection and Clustering,” IEEE Transactions on Communications,
vol. 61, no. 1, pp. 53-62, January 2013.
[206] Y. Zou, Y.-D. Yao, and B. Zheng, “Cognitive transmissions with multiple relays in cognitive radio networks,” IEEE Transactions on
Wireless Communications, vol. 10, no. 2, pp. 648–659, February 2011.
[207] X. Qiao, Z. Tan, S. Xu, and J. Li, “Combined power allocation in cognitive radio-based relay-assisted networks,” IEEE International
Conference on Communications Workshops (ICC 2010), pp. 1–5, May 2010.
[208] Z. Liu, Y. Xu, D. Zhang, and S. Guan, “An efficient power allocation algorithm for relay assisted cognitive radio network,” 2010
International Conference on Wireless Communications and Signal Processing (WCSP), October 2010, pp. 1–5.
[209] O. Oyman, “Opportunistic scheduling and spectrum reuse in relay-based cellular OFDMA networks,” in IEEE Global Telecommu-
nications Conference (GLOBECOM ’07), pp. 3699–3703, November 2007.
[210] R. Wang, V. K. N. Lau, and Y. Cui, “Decentralized fair scheduling in two-hop relay-assisted cognitive OFDMA systems,” in IEEE
Journal on Selected Topics in Signal Processing, vol. 5, no. 1, pp. 171–181, February 2011.
[211] K. Lee and A. Yener, “Outage performance of cognitive wireless relay networks,” in IEEE Global Telecommunications Conference
(GLOBECOM ’06), pp. 1–5., December 2006.
[212] Y. Zou, Y.-D. Yao, and B. Zheng, “Outage probability analysis of cognitive transmissions: Impact of spectrum sensing overhead,” IEEE
Transactions on Wireless Communications, vol. 9, no. 8, pp. 2676–2688, August 2010.
[213] I. F. Akyildiz, D. M. G.-Estevenz, and E. C. Reyes, “The evolution to 4G cellular system: LTE-Advanced,” Physical Communication
(Elsevier), 3 (2010), pp. 217–244.
[214] J. Wang, M. Ghosh, and K. Challapali, “Emerging cognitive radio applications: A survey,” IEEE Communications Magazine, vol. 49,
no. 3, pp. 74–81, March 2011.
[215] W. Lehr and N. Jesuale, “Public safety radios must pool spectrum,” IEEE Communications Magazine, vol. 47, no. 3, pp. 103–109,
March 2009.
[216] E. Leu, B. L. Mark, and M. A. McHenry, “A framework for cognitive WiMAX with frequency agility,” Proceedings of the IEEE, vol.
97, no. 4, pp. 755–773, April 2009.
[217] M. Buddhikot, I. Kennedy, F. Mullany, and H. Viswanathan, “Ultra-broadband femtocells via opportunistic reuse of multi-operator
and multi-service spectrum,” Bell Labs Technical Journal (BLTJ), Special Issue on 4G Networks, February 2009.
[218] G. Gur, S. Bayhan, and F. Alagoz, “Cognitive femtocell networks: An overlay architecture for localized dynamic spectrum access,”
IEEE Wireless Communications, vol. 17, no. 4, pp. 62–70, August 2010.
Page 63 of 66
http://mc.manuscriptcentral.com/wcm
Wireless Communications and Mobile Computing
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
For Peer Review
62
[219] T. Chen, H. Zhang, G. M. Maggio, and I. Chlamtac, “CogMesh: A cluster-based cognitive radio network,” in Proc. of 2nd IEEE
International Symposium on Dynamic Spectrum Access Networks (DySPAN07), pp. 168–178, April 2007.
[220] G. Gur and F. Alagoz, “Green wireless communications via cognitive dimension: An overview,” IEEE Network, vol. 25, no. 2, pp.
50–56, March-April 2011.
[221] M. Nekovee, “A survey of cognitive radio access to TV White Spaces,” in Proc. of International Conference on Ultra Modern
Telecommunications & Workshops (ICUMT), Oct. 2009.
[222] C. Ghosh, S. Roy, and D. Cavalcanti, “Coexistence challenges for heterogeneous cognitive wireless networks in TV white spaces,”
IEEE Wireless Communications, vol. 18, no. 4, August 2011, pp. 22-31.
[223] G. P. Villardi, Y. D. Alemseged, C. Sun, C.-S. Sum, T. H. Nguyen, T. Baykas, and H. Harada, “Enabling coexistence of multiple
cognitive networks in TV white space,” IEEE Wireless Communications, vol. 18, no. 4, pp. 32–40, August 2011.
[224] J. Xiao, F. Ye, T. Tian, and R. Q. Hu, “CR enabled TD-LTE within TV white space: System level performance analysis,” in Proc. of
IEEE Global Telecommunications Conference (GLOBECOM), December 2011.
[225] C.-H. Ko, D. H. Huang, and S.-H. Wu, “Cooperative spectrum sensing in TV White Spaces: When cognitive radio meets cloud,” in
Proc. of IEEE Conference on Computer Communications Workshops (INFOCOM WKSHPS), pp. 672-677, April 2011.
[226] M. A. Rahman, C. Song, and H. Harada, “Development of a TV white space cognitive radio prototype and its spectrum sensing
performance,” in Proc. of International ICST Conference on Cognitive Radio Oriented Wireless Networks and Communications
(CROWNCOM), pp. 231–235, June 2011.
[227] S. Kawade and M. Nekovee, “Can cognitive radio access to TV white spaces support future home networks?,” in Proc. of IEEE
Symposium on New Frontiers in Dynamic Spectrum (DySPAN), April 2010.
[228] A. Zaballos, A. Vallejo, and J. M. Selga, “Heterogeneous communication architecture for the smart grid,” IEEE Network, vol. 25, no.
5, pp. 30–37, September-October 2011.
[229] R. C. Qiu, Z. Hu, Z. Chen, N. Guo, R. Ranganathan, S. Hou, and G. Zheng, “Cognitive radio network for the smart grid: Experimental
system architecture, control algorithms, security, and microgrid testbed,” IEEE Transactions on Smart Grid, vol. 2, no. 4, pp. 724–740,
December 2011.
[230] J. Wang, M. Ghosh, and K. Challapali, “Emerging cognitive radio applications: A survey,” IEEE Communications Magazine, vol. 49,
no. 3, pp. 74–81, March 2011.
[231] A. Ghassemi, S. Bavarian, and L. Lampe, “Cognitive radio for smart grid communications,” in Proc. of IEEE International Conference
on Smart Grid Communications (SmartGridComm), pp. 297–302, October 2010.
[232] R. Yu, Y. Zhang, S. Gjessing, C. Yuen, S. Xie, and M. Guizani, “Cognitive radio based hierarchical communications infrastructure
for smart grid,” IEEE Network, vol. 25, no. 5, pp. 6–14, September-October 2011.
[233] R. Deng, S. Maharjan, X. Cao, J. Chen, Y. Zhang, and S. Gjessing, “Sensing-delay tradeoff for communication in cognitive radio
enabled smart grid,” in Proc. of IEEE International Conference on Smart Grid Communications (SmartGridComm), pp. 155–160,
October 2011.
[234] Q. D. Vo, J.-P, Choi, H. M. Chang, and W. C. Lee, “Green perspective cognitive radio-based M2M communications for smart meters,”
in Proc. of International Conference on Information and Communication Technology Convergence (ICTC), pp. 382–383, November
2010.
[235] C.-K. Yu and K.-C. Chen, “Spectrum map retrieval using cognitive radio network tomography,” in Proc. of IEEE GLOBECOM
Workshops (GC Wkshps), pp. 986–991, December 2011.
[236] K.-D. Lee, S. Kim, and B. Yi, “Throughput comparison of random access methods for M2M service over LTE networks,” in Proc.
of IEEE GLOBECOM Workshops (GC Wkshps), pp. 373–377, December 2011.
Page 64 of 66
http://mc.manuscriptcentral.com/wcm
Wireless Communications and Mobile Computing
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
For Peer Review
63
[237] H. T. Dinh, C. Lee, D. Niyato, and P. Wang, “A survey of mobile cloud computing: Architecture, applications, and approaches,”
Wireless Communications and Mobile Computing, to appear.
[238] V. Namboodiri, “Towards sustainability in portable computing through cloud computing and cognitive radios,” in Proc. of International
Conference on Parallel Processing Workshops (ICPPW), pp. 468–475, September 2010.
[239] K. Nagothu, B. Kelley, M. Jamshidi, and A. Rajaee, “Persistent Net-AMI for microgrid infrastructure using cognitive radio on cloud
data centers,” IEEE Systems Journal, vol. 6, no. 1, pp. 4–15, March 2012.
[240] C.-H. Ko, D. H. Huang, and S.-H. Wu, “Cooperative spectrum sensing in TV white spaces: When cognitive radio meets cloud,” in
Proc. of IEEE Conference on Computer Communications Workshops (INFOCOM WKSHPS), pp. 672–677, April 2011.
[241] F. Ge, H. Lin, A. Khajeh, C. J. Chiang, A. M. Eltawil, C. W. Bostian, W.-chun Feng, and R. Chadha, “Cognitive radio rides on the
cloud,” in Proc. of IEEE Military Communications Conference (MILCOM), pp. 1448–1453, October-November 2010.
[242] C. Wang, C. Jiang, S. Tang, and X.-Y. Li, “Scaling laws of cognitive ad hoc networks over general primary network models,” IEEE
Transactions on Parallel and Distributed Systems, vol. 24, no. 5, pp. 1030-1041, May 2013.
[243] O. N. Gharehshiran, A. Attar, and V. Krishnamurthy, “Collaborative sub-channel allocation in cognitive LTE femto-cells: A cooperative
game-theoretic approach,” IEEE Transactions on Communications, vol. 61, no. 1, pp. 325-334, January 2013.
[244] H. ElSawy and E. Hossain, “Two-tier HetNets with cognitive femtocells: Downlink performance modeling and analysis in a multi-
channel environment,” IEEE Transactions on Mobile Computing, to appear.
[245] D. Feng, C. Jiang, G. Lim, L. J. Cimini Jr., G. Feng, and G. Y. Li, “A survey of energy-efficient wireless communications,” IEEE
Communications Surveys & Tutorials, vol. 15, no. 1, pp. 167-178, First Quarter 2013.
[246] S. Huang, H. Chen, Y. Zhang, and H.-H. Chen, “Sensing-energy tradeoff in cognitive radio networks with relays,” IEEE Systems
Journal, vol. 7, no. 1, pp. 68-76, March 2013.
[247] M. Le Treust, S. Lasaulce, Y. Hayel, and G. He, “Green power control in cognitive wireless networks,” IEEE Transactions on Vehicular
Technology, vol. 62, no. 4, pp. 1741-1754, May 2013.
[248] W. Zhong and J. Wang, “Energy-efficient spectrum sharing strategy selection for cognitive MIMO interference channels,” IEEE
Transactions on Signal Processing, vol. 61, no. 14, pp. 3705-3717, July 2013.
[249] T. Han and N. Ansari, “Auction-based energy-spectrum trading in green cognitive cellular networks,” in Proc. of IEEE International
Conference on Communications (ICC 2013), Budapest, Hungary.
[250] M. C. Oto and O. B. Akan, “Energy-efficient packet size optimization for cognitive radio sensor networks,” IEEE Transactions on
Wireless Communications, vol. 11, no. 4, pp. 1544-1553, April 2012.
[251] S. Park, H. Kim, and D. Hong, “Cognitive radio networks with energy harvesting,” IEEE Transactions on Wireless Communications,
vol. 12, no. 3, pp. 1386-1397, March 2013.
[252] F. Granelli, P. Pawelczak, R. Venkatesha Prasad, K. P. Subbalakshmi, R. Chandramouli, J.A. Hoffmeyer, and S. Berger, “Standardization
and research in cognitive and dynamic spectrum access networks: IEEE SCC 41 efforts and open issues,” IEEE Communications
Magazine, January 2010.
[253] “Overview of the IEEE 802.22 Standard on Wireless Regional Area Networks (WRAN) and Core Technologies,” IEEE 802.22 Working
Group Website, http://ieee802.org/22/.
[254] C. Cordeiro, K. Challapali, D. Birru, and N. S. Shankar, “IEEE 802.22: The first worldwide wireless standard based on cognitive
radios,” in Proc. of First IEEE International Symposium on Dynamic Spectrum Access Networks (DySPAN05), pp. 328–337, 8-11
November 2005.
[255] C. Stevenson, G. Chouinard, Z. Lei, W. Hu, S. Shellhammer, and W. Caldwell, “IEEE 802.22: The first cognitive radio wireless
regional area network standard,” IEEE Communications Magazine, vol. 47, no. 1, pp. 130–138, January 2009.
Page 65 of 66
http://mc.manuscriptcentral.com/wcm
Wireless Communications and Mobile Computing
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
For Peer Review
64
[256] M. Murroni et al., “IEEE 1900.6 spectrum sensing interfaces and data structures for dynamic spectrum access and other advanced
radio communication systems standard: Technical aspects and future outlook,” IEEE Communications Magazine, vol. 49, no. 12, pp.
118–127, December 2011.
[257] IEEE, “Standard Definitions and concepts for Spectrum Management and Advanced Radio Technologies,” P1900.1 drafts std, v.031,
June 2007.
[258] O. Holland et al., “Development of a Radio Enabler for Reconfiguration Management within the IEEE P1900.4 Working Group,” in
Proc. of IEEE DySPAN 2007, Dublin, Ireland, April 2007.
[259] S. Buljore, V. Merat, H. Harada, S. Filin, and P. Houze, “IEEE P1900.4 system overview on architecture and enablers for optimised radio
and spectrum resource usage,” in Proc. of IEEE symposium on New Frontiers in Dynamic Spectrum Access Networks (DySPAN’08),
October 2008.
[260] SDR Forum, “SDRF-06-P-0009-V1.0.0,” http://www.sdrforum.org/pages/documentLibrary/documents/SDRF-06-P-0009-V1 0 0
CRWG Defs.pdf
[261] ETSI, “ETSI TR 102 838: Reconfigurable radio systems (RRS); Summary of feasibility studies and potential standardization topics,”
http://www.etsi.org, 2009-10.
[262] ITU-R Report 2063, “The Impact of Software Defined Radio on IMT-2000, the Future Development of IMT-2000 and Systems Beyond
IMT-2000.”
[263] ITU-R Report M.2064, “Software-Defined Radio in the Land Mobile Service.”
Page 66 of 66
http://mc.manuscriptcentral.com/wcm
Wireless Communications and Mobile Computing
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960