SPECTRUM SENSING IN COGNITIVE RADIO NETWORKS: QOS CONSIDERATIONS

David C. Wyld et al. (Eds) : NETCOM, NCS, WiMoNe, CSEIT, SPM - 2015
pp. 09–19, 2015. © CS & IT-CSCP 2015 DOI : 10.5121/csit.2015.51602
SPECTRUM SENSING IN COGNITIVE
RADIO NETWORKS: QOS
CONSIDERATIONS
Nabil Giweli, Seyed Shahrestani and Hon Cheung
School of Computing, Engineering and Mathematics,
Western Sydney University, Sydney, Australia
ngiweli@scem.westernsydney.edu.au,
s.shahrestani@westernsydney.edu.au,
h.cheung@westernsydney.edu.au
ABSTRACT
The rapidly growing number of wireless communication devices has led to massive increases in
radio traffic density, resulting in a noticeable shortage of available spectrum. To address this
shortage, the Cognitive Radio (CR) technology offers promising solutions that aim to improve
the spectrum utilization. The operation of CR relies on detecting the so-called spectrum holes,
the frequency bands that remain unoccupied by their licensed operators. The unlicensed users
are then allowed to communicate using these spectrum holes. As such, the performance of CR is
highly dependent on the employed spectrum sensing methods. Several sensing methods are
already available. However, no individual method can accommodate all potential CR operation
scenarios. Hence, it is fair to ascertain that the performance of a CR device can be improved if
it is capable of supporting several sensing methods. It should obviously also be able to select the
most suitable method. In this paper, several spectrum sensing methods are compared and
analyzed, aiming to identify their advantages and shortcomings in different CR operating
conditions. Furthermore, it identifies the features that need to be considered while selecting a
suitable sensing method from the catalog of available methods.
KEYWORDS
Cognitive Radio; Spectrum Sensing; Qos
1. INTRODUCTION
In general, the Radio Frequency Spectrum (RFS) is statically divided into licensed and unlicensed
bands. While the use of the former is restricted to authorized operators, the unlicensed bands are
available for use by the public, only subject to transmission constraints [1]. As such, the
unlicensed bands may get heavily congested. On the other hand, several studies and
measurements conducted around the world have indicated that the licensed RFS bands can be
underutilized [2].
From a technical perspective, the Cognitive Radio (CR) concept is a promising technology to
achieve an efficient utilization of RFS. The concepts for CR models were introduced in 1999 by
Joseph Mitola [3]. In these models, the operators licensed to use some particular frequency bands
are considered as the Primary Users (PUs). Whereas, the unlicensed participants are referred to as
the Secondary Users (SUs). The CR model is based on the realization that a PU may not fully
utilize its licensed bands, leaving parts of its spectrum unoccupied. These unoccupied white
spaces, or holes, relate to use, or more correctly the lack of use, in terms of frequency, time, or

10 Computer Science & Information Technology (CS & IT)
space and location. An SU can utilize these holes in addition to the unlicensed bands that it may
typically use.
To achieve their objectives, CR systems are dependent on the execution of a sequence of several
functions, the so-called CR cycle. A typical CR cycle was proposed by Mitola [3]. This is
illustrated in Figure 1.
Figure 1. Basic CR cycle
The main functions of this CR cycle are spectrum sensing, spectrum decision, spectrum sharing,
and spectrum mobility. More specifically, an SU should be able to perform the following [4]:
• Spectrum sensing: sense the surrounding RFS to determine spectrum holes and to detect
the presence of the relevant PU.
• Spectrum decision: analyze and decide which spectrum hole is the most suitable for
satisfying the application requirements.
• Spectrum sharing: share the available spectrum holes with other SUs as fairly as possible.
• Spectrum mobility: seamlessly switch to another suitable spectrum hole to avoid
interference with a detected PU that may wish to start using its licensed spectrum.
Detecting the presence of a PU, or more precisely finding out whether the PU is using its
allocated spectrum or not, is an essential task for a CR device. On one hand, this fundamental task
requires improving sensing accuracy by avoiding false positive results while detecting the
presence of a PU. On the other, the employed sensing technique should achieve a high detecting
probability of the available spectrum holes. The nature of the electromagnetic signals makes
accurate sensing a complicated process. More specifically, the Signal to Noise Ratio (SNR), the
multipath fading of the PU signals, and the changing levels of noise can significantly affect the
sensing accuracy [5, 6]. Moreover, imperfect spectrum sensing can result in increased
transmission error rates, for both the PU and the SUs [7]. Such errors may contribute to the
degradation of the quality of the services provided by a PU and SUs. Noticeably, any QoS
degradation that can be attributed to the CR technology can potentially harm the progress of the
CR-based solutions. In this paper, the main features and limitations of the prevalent spectrum
sensing methods are examined. Furthermore, the key aspects that should be involved in selecting
the appropriate sensing method are highlighted and discussed.
The remainder of this paper is organized as follows. Section II presents the background and
motivation for this work. The effects of the sensing operation on the QoS of the applications
running over CR networks are described in Section III. Several sensing approaches are discussed
and evaluated in Section IV. Factors that may help in selecting the proper sensing techniques are
outlined in Section V. The last section gives the conclusions and points to the potential future
expansion of the reported work.

Computer Science & Information Technology (CS & IT) 11
2. MOTIVATIONS
Most of the previous reviews of spectrum sensing techniques are mainly focused on the operation,
accuracies, complexities, and implementation issues [8-11]. For instance, the relation between the
sensing accuracy, and the speed, i.e., sensing time, and frequency, i.e., repeating the sensing, are
the primary focus of the authors in [8]. They aim to achieve an optimal spectrum sensing
performance with the capability of flexible tuning between the speed and frequency. However,
they find that the available state-of-art sensing technologies do not offer a possible trade-off
between complexity and accuracy. In contrast, other reviewers consider the characteristics of the
PU signal as the main factor for selecting a proper sensing method [9]. Nevertheless, other factors
should be considered for more adaptive sensing and improved performance.
In general, the dominant approach is about how to find an optimal sensing method for all possible
CR operation requirements. However, none of the proposed sensing methods is suitable for all
possible sensing situations, conditions and technologies of CR systems. This study shifts the
focus to another approach where a CR device supports a range of various sensing methods. Thus,
the proper sensing technique can be selected based on the real-time requirements. This approach
implies the need for a real-time mechanism to select the most suitable sensing method. In this
paper, various sensing methods are studied toward finding the relevant selection criteria that
should be considered when designing such as real-time selection mechanism.
3. SENSING OPERATION IMPACT ON APPLICATIONS’ QOS IN CR
NETWORKS
As shown in Figure 2, the operation of a CR can be divided into repeated cycles of the sensing
period. The sensing period T effectively represents the time interval where sensing is repeated. It
also represents the communication frame in CR. During the sensing time t, a CR device obtains
information from its environment. After the sensing time, the CR device can decide to transmit
data on the same channel or in a new vacant channel, i.e., a spectrum hole.
Figure 2. Simple structure of CR frames based on sensing operation
The decision taken by the CR device is based on the sensing outcome, i.e., the presence or
absence of the PU. The transmission starts after the sensing time until the next sensing period,
also called the CR communication frame. The transmission time (T-t) depends on the sensing
time t and the frame time T that is based on the design of how frequent the sensing will be
conducted. Thus, sensing frequency is 1/T. The sensing time t and the frame time T can be
designed to be fixed for all frames or could be designed to vary based on the design goals [12].
Typically, the sensing operation should be limited and less frequent as much as possible without
affecting the sensing accuracy [13].
Increasing the sensing time t and conducting the sensing more frequently, i.e. decreasing T, lead
to an increase in the probability of correct detection of the PU’s presence. In turn, this leads to
more protection to the PU from interference by CR users and more utilization of the spectrum. On
the other hand, this leads to less data transmission rate and hence to QoS degradation for SUs.

The degradation can be measured by several parameters such as throughput, delay and MAC
layer process overhead [14]. Therefore, designing the sensing time and frequency of sensing
operation should take into account the trade-off between protecting the PU’s QoS and improving
the QoS of SUs.
4. SENSING METHODS
The main challenge facing the sensing methods is how to improve the spectrum sensing
performance by mainly increasing the positive detection probability and decreasing the false
detection probability. A sensing technique with a higher positive detection probability provides
more protection to PU. A CR user with a lower probability of false detection of the presence of
the PU has more chance to use the available spectrum holes. Therefore, the user has more chance
of achieving a higher throughput on the CR network. The design of a sensing technique is
constrained by an acceptable level of false detection [15]. Additionally, improving sensing
performance is challenged by a range of trade-offs and various constraints such as application
requirements, hardware capability, complexity and required infrastructure [16].
In general, a sensing method that uses surrounding RFS information collected by the CR device
only is called a local sensing. If the SUs do not exchange their surrounding RFS information
gathered by local sensing, then this sensing is referred to as non-cooperative sensing. In this
paper, the sensing methods are classified mainly into three categories: methods with no prior
information required, based on prior information and based on SUs cooperation.
4.1. No Prior Information Required (Blind Sensing)
No prior information about the PUs’ signal is necessary for the sensing methods under this
category. However, prior information about the noise power of the targeted spectrum may be
required for better performance. Otherwise, a reasonable estimation of the noise power is used
instead. Two well-known blind sensing methods are energy detection and covariance-based
detection.
4.1.1. Energy Detection
Also known as radiometry or periodogram, energy detection is the most common method for
spectrum sensing because of its low implementation complexity and computational overhead [5].
In this method, the energy detector is used to detect a narrowband spectrum and then the observed
signal energy level is compared with a predefined threshold. Thus, the channel is occupied by the
PU if the detected signal energy is over the threshold. Otherwise, it is considered unoccupied, i.e.,
a spectrum hole. Because of this simplicity, this technique requires the shortest sensing duration t
per frame compared to other common sensing technologies [17].
Generalizing the use of this method faces several challenges as a consequence of its simplicity.
Firstly, selection of the threshold used for detection is an issue when the channel noise level is
unknown or uncertain over time [18]. Secondly, under low SNR, it is hard to differentiate
between modulated signals, including signals of other SUs, noise, and interference, resulting in
poor detection performance [5]. Lastly, an energy detector is ineffective in detecting spread
spectrum signals [19].
4.1.2. Covariance-based Detection
This method is based on comparing the covariance of the detected signal and the covariance of
the noise where statistical covariance matrices of signal and noise are usually different [20]. The
main improvement of this method is to overcome the energy detection shortcoming. In particular,
it can distinguish between signal and noise in a low SNR, and without any prior information
about the PU’s signal and channel noise. This detection improvement is achieved at the expense

of adding a computational overhead in computing the covariance matrix of the received signal
samples [11]. In addition to increasing complexity, other drawbacks of the energy detection are
still present in the covariance-based detection.
These sensing methods work with no prior information about the PU signals. They have a limited
performance particularly for spread spectrum and in situations where other SUs are sharing the
spectrum. Research is ongoing to improve the blind sensing approach in terms of performance
and required sensing time, such as in [21, 22].
4.2. Prior Information Required
Methods belonging to this category rely on partial or full information about the PU’s transmission
signal to be able to differentiate it from other signals and noise.
4.2.1. Cyclostationarity Feature Detection
This method is based on distinguishing the PU signal from noise, interference, and other signals
by identifying its cyclostationarity features [23, 24]. These cyclostationarity features are
associated with the signal modulation type, carrier frequency, and data rate. Hence, the CR device
needs sufficient prior information about these unique characteristics of the PU signal. Based on
this information, it can perform a cyclostationarity analysis on the detected signal to identify
matched features [9]. For this method to perform better than the energy detection method, an
adequate number of real-time sample sets in the frequency domain need to be collected. As a
consequence, better performance accrues more complexity and sensing time at the expense of the
available throughput [9].
4.2.2. Correlation Detection
Sensing based on correlation is also known as waveform-based sensing or coherent sensing. In
this method, the expected correlation or coherence between signal samples is identified to detect
the PU signal based on previous knowledge about its waveform patterns [9]. The accuracy of the
sensing increases when the length of the known signal pattern of the PU is increased [25]. The
main drawback of this method is related to the large amount of information required for signal
patterns of the PUs to achieve a high performance that is not practical for all CR systems.
4.2.3. Radio Identification Based Sensing
This method is based on having apriori information about the transmission technologies used by
the PU. In the radio identification stage of the method, several features of the received signal are
exploited and then classified to determine if the signal demonstrates the PU signal technology
[26]. Fundamentally, the feature extraction and classification techniques are used in the context of
European Transparent Ubiquitous Terminal (TRUST) project [27]. For collecting the signal
features, the radio identification method may use one of the known sensing techniques, such as
the energy detection method [9]. The radio identification improves the accuracy of the energy
detection to some extent with complexity implication. The achieved precision is dependent on the
signal features and classification techniques used to identify the presence of the PU.
4.2.4. Matched Filtering
The matched filtering method achieves a higher detection probability in a short detection time,
compared to other methods that are similarly based on prior information [28, 29]. Hence, under
this classification, this method is considered as the best sensing method. The collected signal is
passed through a filter that will amplify the possible PU signal and attenuate any noise signal. The
filter makes the detection of the presence of the PU signal more accurate [29]. The filter, which
is known as a matched filter, has to be tuned based on some features of the PU signal. These

characteristics include the required bandwidth, operating frequency, the modulation used and
frame format [9]. One of the disadvantages of this method is in implementation where different
PUs signal types require different dedicated hardware receivers. This requirement makes the
method impractical to implement and also leads to higher power consumptions if the method is
implemented based on current hardware technologies.
Figure 3 shows a comparison between non-cooperative sensing methods, based on accuracy and
complexity metrics. Table 1 shows more comparison factors between local sensing methods.
Figure 3. Sensing method complexity versus accuracy
Table 1. Comparison between local sensing methods
4.3. Based on SU Cooperation
The main principle of this approach is that SUs share their local sensed information of the
spectrum. The use of sensed information from all SUs can produce a more accurate sensing
outcome than relying solely on local sensing. The hidden transmitter problem is an example of
the issues that may prohibit a CR from detecting the presence of a PU. The cause of this problem
is the fading and shadowing of the signals from a PU, although it is within the transmission range
of the CR [9]. However, when cooperated SUs are spatially distributed, it helps to overcome the
hidden PU problem and other limitations of local sensing [30]. Sensing cooperation can also
reduce the local sensing cost, e.g., sensing time duration and energy consumption while
maintaining sensing quality by scheduling the sensing operation among cooperative SUs [31].

The sensing method used by an individual SU can be based on one of the sensing methods for
local sensing, such as energy detection and cyclostationarity feature detection [10].
In some environments, cooperative sensing may lose its advantages as far as an individual SU is
concerned. For instance, increasing the local sensing frequency in individual high mobility SUs
is more efficient, in terms of sensing accuracy and overhead, than to cooperate with other SUs
[19]. In cooperative sensing, the improvement of sensing is more noticeable when the number of
cooperative SUs is increased. However, involvement of more SUs will increase the cooperation
overhead in terms of the amount of data exchange and the time required for the exchange [32].
The cooperative approaches can only be used when SUs are able and willing to collaborate. Also,
a SU may not always find other cooperative SUs within its transmission range. Therefore, the CR
devices should not solely rely on cooperative sensing approaches. They should be able to use a
fitting local sensing method and resort to cooperative sensing, only when an enhanced
performance is possible.
5. FACTORS FOR SELECTING THE FITTING SENSING METHOD
Selecting the best sensing method for a particular cognitive radio operation condition depends on
several factors. Based on the discussions in previous sections, notable factors are summarized
below:
5.1. CR Device Capability
A CR device designed with limited hardware resources and power capacities will not be able to
support a wider range of sensing methods. Some methods require sophisticated hardware
components and higher power consumption, e.g. the matched filter method, compared to simple
ones such as the energy detection method. An ideal CR device should be able to be reconfigured
on-the-fly to support a broad range of sensing methods. In practice, a CR device’s actual
capability will limit the range of sensing methods that can be supported.
5.2. Qos Required for Applications Running on the CR Device
The QoS requirements differ based on the applications running on a CR device. The sensing
delay and transmission throughput vary from one sensing method to another within the same
conditions. As a result, the sensing operation used on a CR device has a direct impact on the QoS
of an application running on the device, mainly in terms of the throughput and delay. As sensing
is a repetitive operation, a CR device should be able to select a proper sensing method with the
least impact on the QoS of the running application. Other operational requirements must also be
taken into account. For example, the PU protection should have a higher priority than the QoS
requirements of a CR user.
5.3. Apriori Information
The extent of information available about the characteristics of the PUs and the communications
media is a major factor influencing the selection of a proper sensing method. For instance,
insufficient information about the PU signals, excludes the use of matched filter method.
The CR device should be able to change the sensing method based on the information that
becomes available about the PU signal or the SNR of the targeted spectrum by sensing.

5.4. Level of Protection Required for PU
The selection of the sensing method must be considered with regard to the degree of protection
necessary for the PU. They may vary depending on available frequency bands and types of
services. For instance, analog TV service is more robust against interference than digital TV
service [16]. Hence, a sensing method that provides less protection, i.e., lower PU detection
probability, should only be used when the PU is more tolerant of interference such as in analog
TV services.
5.5. The CR Network Mode and Capability
The network mode and capability are important factors to CR systems to make a decision
between cooperative and non-cooperative sensing approaches. In CR networks with infrastructure
and centralized topology, a method based on cooperative sensing is more suitable than that based
on local sensing only. Hence, the capability of such a CR network depends on how much
management ability can provide for white space determination to its CR devices. Furthermore, the
capacity of a CR network relies on how much information the network can gather and provide to
its CR users about the PU signals and the ambient spectrum.
6. CONCLUSIONS
The work reported in this paper asserts that none of the available spectrum sensing techniques can
achieve perfect solutions for all potential CR operating conditions. Therefore, to improve the
performance of CR systems, the relevant devices must be capable of utilizing a catalog of sensing
methods. The selection of the most suitable method from the catalog, which is an obvious
necessity, is based on a number of factors that have also been identified and discussed in this
paper. Our future works will focus on more exhaustive evaluations of these factors and how their
fine-tunings can contribute to an improved CR performance.
REFERENCES
[1] L. Cui and M. B. Weiss, "Can unlicensed bands be used by unlicensed usage," Paper for the 41st
annual TPRC, September, pp. 27-29, 2013.
[2] K. Patil, R. Prasad, and K. Skouby, "A Survey of Worldwide Spectrum Occupancy Measurement
Campaigns for Cognitive Radio," in Devices and Communications (ICDeCom), 2011 International
Conference on, 2011, pp. 1-5.
[3] J. Mitola and G. Q. Maguire Jr, "Cognitive radio: making software radios more personal," Personal
Communications, IEEE, vol. 6, pp. 13-18, 1999.
[4] J. Marinho and E. Monteiro, "Cognitive radio: survey on communication protocols, spectrum decision
issues, and future research directions," Wireless networks, vol. 18, pp. 147-164, 2012/02/01 2012.
[5] Y. Zeng, Y.-C. Liang, A. T. Hoang, and R. Zhang, "A review on spectrum sensing for cognitive radio:
challenges and solutions," EURASIP J. Adv. Signal Process, vol. 2010, pp. 2-2, 2010.
[6] A. F. Molisch, L. J. Greenstein, and M. Shafi, "Propagation issues for cognitive radio," Proceedings
of the IEEE, vol. 97, pp. 787-804, 2009.
[7] S. Haddadi, H. Saeedi, and K. Navaie, "Channel coding adoption versus increasing sensing time in
secondary service to manage the effect of imperfect spectrum sensing in cognitive radio networks," in
Communication and Information Theory (IWCIT), 2013 Iran Workshop on, 2013, pp. 1-5.

[8] D. Ariananda, M. Lakshmanan, and H. Nikookar, "A survey on spectrum sensing techniques for
cognitive radio," in Cognitive Radio and Advanced Spectrum Management, 2009. CogART 2009.
Second International Workshop on, 2009, pp. 74-79.
[9] T. Yucek and H. Arslan, "A survey of spectrum sensing algorithms for cognitive radio applications,"
Communications Surveys & Tutorials, IEEE, vol. 11, pp. 116-130, 2009.
[10] I. F. Akyildiz, B. F. Lo, and R. Balakrishnan, "Cooperative spectrum sensing in cognitive radio
networks: A survey," Physical Communication, vol. 4, pp. 40-62, 2011.
[11] D. B. Rawat and G. Yan, "Spectrum sensing methods and dynamic spectrum sharing in cognitive
radio networks: A survey," International Journal of Research and Reviews in Wireless Sensor
Networks, vol. 1, pp. 1-13, 2011.
[12] P. Yiyang, L. Ying-Chang, K. C. Teh, and L. Kwok Hung, "How much time is needed for wideband
spectrum sensing?," Wireless Communications, IEEE Transactions on, vol. 8, pp. 5466-5471, 2009.
[13] C. Kae Won, "Adaptive Sensing Technique to Maximize Spectrum Utilization in Cognitive Radio,"
Vehicular Technology, IEEE Transactions on, vol. 59, pp. 992-998, 2010.
[14] H. Xin-Lin, W. Gang, H. Fei, and S. Kumar, "The Impact of Spectrum Sensing Frequency and
Packet-Loading Scheme on Multimedia Transmission Over Cognitive Radio Networks," Multimedia,
IEEE Transactions on, vol. 13, pp. 748-761, 2011.
[15] L. Ying-Chang, Z. Yonghong, E. C. Y. Peh, and H. Anh Tuan, "Sensing-Throughput Tradeoff for
Cognitive Radio Networks," Wireless Communications, IEEE Transactions on, vol. 7, pp. 1326-1337,
2008.
[16] A. Ghasemi and E. S. Sousa, "Spectrum sensing in cognitive radio networks: requirements, challenges
and design trade-offs," Communications Magazine, IEEE, vol. 46, pp. 32-39, 2008.
[17] H. Kim and K. G. Shin, "In-band spectrum sensing in cognitive radio networks: energy detection or
feature detection?," in Proceedings of the 14th ACM international conference on Mobile computing
and networking, 2008, pp. 14-25.
[18] R. Tandra and A. Sahai, "SNR Walls for Signal Detection," Selected Topics in Signal Processing,
IEEE Journal of, vol. 2, pp. 4-17, 2008.
[19] D. Cabric, S. M. Mishra, and R. W. Brodersen, "Implementation issues in spectrum sensing for
cognitive radios," in Signals, Systems and Computers, 2004. Conference Record of the Thirty-Eighth
Asilomar Conference on, 2004, pp. 772-776 Vol.1.
[20] Y. Zeng and Y.-C. Liang, "Spectrum-sensing algorithms for cognitive radio based on statistical
covariances," Vehicular Technology, IEEE Transactions on, vol. 58, pp. 1804-1815, 2009.
[21] Y. Li, S. Shen, and Q. Wang, "A Blind Detection Algorithm Utilizing Statistical Covariance in
Cognitive Radio," International Journal of Computer Science Issues (IJCSI), vol. 9, pp. 7-12, 2012.
[22] K. S. Kumar, R. Saravanan, and R. Muthaiah, "Cognitive Radio Spectrum Sensing Algorithms based
on Eigenvalue and Covariance methods," International Journal of Engineering & Technology (0975-
4024), vol. 5, 2013.
[23] J. Lundén, V. Koivunen, A. Huttunen, and H. V. Poor, "Spectrum sensing in cognitive radios based
on multiple cyclic frequencies," in Cognitive Radio Oriented Wireless Networks and
Communications, 2007. CrownCom 2007. 2nd International Conference on, 2007, pp. 37-43.
[24] S. Enserink and D. Cochran, "A cyclostationary feature detector," in Signals, Systems and Computers,
1994. 1994 Conference Record of the Twenty-Eighth Asilomar Conference on, 1994, pp. 806-810.

[25] H. Tang, "Some physical layer issues of wide-band cognitive radio systems," in New frontiers in
dynamic spectrum access networks, 2005. DySPAN 2005. 2005 first IEEE international symposium
on, 2005, pp. 151-159.
[26] T. Yucek and H. Arslan, "Spectrum characterization for opportunistic cognitive radio systems," in
Military Communications Conference, 2006. MILCOM 2006. IEEE, 2006, pp. 1-6.
[27] T. Farnham, G. Clemo, R. Haines, E. Seidel, A. Benamar, S. Billington, et al., "Ist-trust: A
perspective on the reconfiguration of future mobile terminals using software download," in Personal,
Indoor and Mobile Radio Communications, 2000. PIMRC 2000. The 11th IEEE International
Symposium on, 2000, pp. 1054-1059.
[28] S. Shobana, R. Saravanan, and R. Muthaiah, "Optimal Spectrum Sensing Approach on Cognitive
Radio Systems," 2013.
[29] S. Shobana, R. Saravanan, and R. Muthaiah, "Matched Filter Based Spectrum Sensing on Cognitive
Radio for OFDM WLANs," International Journal of Engineering and Technology (IJET), vol. 5,
2013.
[30] W. Chien-Min, S. Hui-Kai, L. Maw-Lin, L. Yi-Ching, and L. Chih-Pin, "Cooperative Power and
Contention Control MAC Protocol in Multichannel Cognitive Radio Ad Hoc Networks," in
Innovative Mobile and Internet Services in Ubiquitous Computing (IMIS), 2014 Eighth International
Conference on, 2014, pp. 305-309.
[31] X. Dongyue, E. Ekici, and M. C. Vuran, "Cooperative Spectrum Sensing in Cognitive Radio
Networks Using Multidimensional Correlations," Wireless Communications, IEEE Transactions on,
vol. 13, pp. 1832-1843, 2014.
[32] Z. Yulong, Y. Yu-Dong, and Z. Baoyu, "Cooperative relay techniques for cognitive radio systems:
Spectrum sensing and secondary user transmissions," Communications Magazine, IEEE, vol. 50, pp.
98-103, 2012.
AUTHORS
Nabil Giweli received the B.Sc. degree in Communication Engineering from
Tripoli University, Libya, in 1997, the Master degree in Information and
Communication Technology (with the dean medal award) from the Western
Sydney University in 2011, and another M.Sc. form the same university in
Cloud Security in 2013. Currently, he is a Ph.D. candidate and a casual teacher
at the School of Computing, Engineering and Mathematics, Western Sydney
University, Australia. His current research area is in Cognitive Radio
Technologies.
Seyed Shahrestani completed his PhD degree in Electrical and Information
Engineering at the University of Sydney. He joined Western Sydney
University (Western) in 1999, where he is currently a Senior Lecturer. He is
also the head of the Networking, Security and Cloud Research (NSCR) group
at Western. His main teaching and research interests include: computer
networking, management and security of networked systems, analysis, control
and management of complex systems, artificial intelligence applications, and
health ICT. He is also highly active in higher degree research training
supervision, with successful results.

Dr Hon Cheung graduated from The University of Western Australia in 1984
with First Class Honours in Electrical Engineering. He received his PhD degree
from the same university in 1988. He was a lecturer in the Department of
Electronic Engineering, Hong Kong Polytechnic from 1988 to 1990. From 1990
to 1999, he was a lecturer in Computer Engineering at Edith Cowan University,
Western Australia. He has been a senior lecturer in Computing at Western
Sydney University since 2000. Dr Cheung has research experience in a number
of areas, including conventional methods in artificial intelligence, fuzzy sets,
artificial neural networks, digital signal processing, image processing, network security and forensics, and
communications and networking. In the area of teaching, Dr Cheung has experience in development and
delivery of a relative large number of subjects in computer science, electrical and electronic engineering,
computer engineering and networking.

SPECTRUM SENSING IN COGNITIVE RADIO NETWORKS: QOS CONSIDERATIONS

More Related Content

What's hot (19)

Similar to SPECTRUM SENSING IN COGNITIVE RADIO NETWORKS: QOS CONSIDERATIONS (20)

More from cscpconf (20)

Recently uploaded (20)

SPECTRUM SENSING IN COGNITIVE RADIO NETWORKS: QOS CONSIDERATIONS