study on energy detection-based cooperative sensing in ......gaussian noise. a denotes the...

Study on Energy Detection-based Cooperative Sensing in Cognitive Radio Networks

Rania Mokhtar

Department of Electronic Engineering, Faculty of Engineering Sudan University of Science and Technology, Khartoum, Sudan

Email: [email protected]

1Rashid Saeed, 2*Raed Alsaqour, and 3Yahia Abdallah 1Faculty of Engineering, Sudan University of Science and Technology (SUST), Khartoum, Sudan

2Faculty of Information Science and Technology, University Kebangsaan Malaysia, Bangi, 43600, Selangor, Malaysia 3

Email:

Faculty of Computer Science and Information Technology, Sudan University of Science and Technology, Khartoum, Sudan

[email protected], [email protected], 3

[email protected]

Abstract—Cognitive radio (CR) technology aims to achieve efficient radio spectrum utilization based on overlay spectrum sharing. Therefore, one of the main requirements of CR systems is the capability to detect and sense the presence of primary transmissions. However, sensing is affected by the behavior of the wireless channel, i.e., hidden node, interference, shadowing, and fading, which may result in wrong detection decisions. Consequently, CRs may introduce harmful interference to primary radios. Cooperative spectrum sensing can improve detection decisions by obtaining sensing information from different sources, which has been recently proposed to overcome this problem. This paper reviews cooperative sensing aspects, approaches, architecture, as well as problematic aspect and drawbacks of the control channel and associated fusion methods. Index Terms—Cognitive radio; spectrum sensing; cooperative sensing; data fusion

I. INTRODUCTION

The basic problem concerning spectrum sensing is the signal detection within a noisy measure. The performance of local spectrum sensing techniques is limited by the received signal strength. Spectrum sensing in a cognitive radio (CR) node (local sensing) is associated with several challenges [1]. The channel state information among a primary transmitter, a receiver, and a receiver location is unknown to a CR. The hidden node problem can arise when a blockage exists between the TV whitespace device and a TV station, but no blockage exists between the TV station and a TV receiver antenna and between the unlicensed device and the same TV receiver antenna. In such a case, a CR may not detect the presence of a TV signal and can start using an occupied channel, causing harmful interference to the TV receiver. This problem increases the requirement in the CR sensitivity to a level that outperforms primary user (PU) receivers to the extent that it becomes capable of detecting weak signals. Local node sensing may achieve an acceptable sensing result only after an exceedingly long sensing time. Considering this limitation in local spectrum sensing, the cooperation

among CRs is introduced for better accuracy of PU detection. Cooperation may solve of shadowing and hidden node problem [2], as shown in Fig. 1. Cooperation is considered as the key method for the realization of a CR.

Figure 1. Cooperative detection

Sensing can be improved by incorporating more

observations from other CRs at various locations. Network cooperation can improve sensing performance compared with local sensing, whereby cooperative sensing enhances the detection and reduces the probability of interference to a PU. Cooperative sensing is based on the fact that summing signals from two CRs can improve detection reliability by increasing the signal-to-noise ratio (SNR) if signals are correlated [3]. Cooperative spectrum sensing exploits the broadcast nature and spatial diversity of the channel. In cooperative sensing, CR receivers may estimate channel variations resulting from fading, noise, and interference by relaying messages to one another. These messages propagate redundant signals over multiple paths in the network [4].

Realization of CR and white space standard [3] employs a set of applications, such as relay channels, distributed antenna arrays, localization methods, data exchange and fusion methods as well as collaborative mapping. Cooperative sensing can be implemented in different ways based on the cooperation method and network architecture. Cooperative sensing methods

JOURNAL OF NETWORKS, VOL. 8, NO. 6, JUNE 2013 1255

© 2013 ACADEMY PUBLISHERdoi:10.4304/jnw.8.6.1255-1261

incorporate a set of cooperation technologies including network architecture, sensing method, data fusion algorithm, sensor selection, and data exchange protocol that characterizes method performance. This paper reviews cooperative sensing aspects and approaches as well as the problematic aspect of the control channel and associated fusion methods. In this paper, we also propose a new cooperative sensing framework. Fig. 2 identifies the direction of research on collaborative sensing. The chart shows cooperative technologies employed in spectrum sensing and highlights another related cooperative method in the area.

In this paper, we performed a comparative study on different cooperative sensing methods and their network architectures, problem considerations, and drawbacks. This study provides guidance on designing a sensing method for a cooperative sensing network by considering the introduced problems and drawbacks.

The remainder of the paper is organized as follows: Section II is reviews sensing-based on cooperative transmission. Section III discusses the cooperative sensing framework. Section IV presents the cooperative sensing architecture and comparison. We conclude the paper in Section V.

Figure 2. Research direction in collaborative sensing

II. COOPERATIVE TRANSMISSION-BASED SENSING

Cooperative sensing can exploit both cognitive relay and cooperative transmission in the detection of the PU. CR can forward/relay PU traffic toward the destination. This kind of cooperation is known as cognitive relay. By relaying PU traffic, the primary network gains in terms of throughput and diversity to provide more transmission

opportunities. On the other hand, relaying PU traffic can enhance PU detection.

Figure 3. Two CR cooperative scheme

Cognitive relay uses a fixed relay protocol, in which a cognitive user encodes and decodes or amplifies and forwards the primary signal. Cognitive relay can be employed by users located at the boundary of decodablity to enhance the detection of weak signals [5]. We consider two cognitive nodes, using a fixed time slot mode time division multiple access (TDMA) to send messages to a common receiver [6].

Meanwhile, if a PU starts using the frequency, then the presence of this PU should be detected as soon as possible so that the band can be vacated immediately. If one of the cognitive nodes is located at the decodablity border as shown in Fig. 3, then it will receive an extremely weak signal, thus taking more time to sense the PU. However, the time taken to detect the weak signal is reduced if cooperative transmission is employed as shown in Fig. 4, where cognitive relay employ an amplify-and-forward protocol for the PU signal.

In time slot 1T , the received PU signal at the cognitive relay is

wahhy p ++= 122θ (1)

Figure 4. Relay transmission protocol in two users cooperative

where pih denotes the instant channel gain between CRi

and the PU, and 12h denotes the instant channel gain between the two CR nodes, and w is the additive white

1256 JOURNAL OF NETWORKS, VOL. 8, NO. 6, JUNE 2013

© 2013 ACADEMY PUBLISHER

Gaussian noise. a denotes the transmitted signal from cognitive node CR1, such that 0}{ =aE and θ denotes the primary absence/presence pointer; 1=θ denotes the existence of the PU, and 0=θ implies an idle channel. In slot 2T , cognitive relay receives its sent message as the received node is relaying the message again. The signal received by cognitive relay is

whhyy p ++= 112 θϕ

whwahhh pp ++++= 112212 )( θθϕ (2)

where 1ph is the instant channel gain between CR1 and PU, w is white Gaussian noise, and ϕ is the scaling factor [8] used by CR2 to relay the information of CR1 to the common receiver. Equation (2) can be written as:

WHY += θ (3)

where 2121 pp hhhH ϕ+= and whwW 12ϕ+= . Given this equation, the detection hypothesis can be stated as

1:1 =θH or 0:0 =θH . The challenge with this scenario is that cognitive relay should have no power constraint or should adjust its transmit power. This condition determines the upper bound of cooperation performance.

III. COOPERATIVE SENSING FRAMEWORK

The distributed cooperative sensing (DCS) framework involves making observations of the radio frequency (RF) usage, sensing management, cooperative management, and sensing policies. Fig. 5 illustrates the sensing functions in the physical, MAC, and network layers. A variety of methods and techniques are implemented and mapped to achieve the sensing stack functionalities. However, the interference temperature and wideband sensing models are not implemented in the sensing stack.

Figure 5. Sensing stack functionalities

IV. CENTRALIZED COOPERATIVE ARCHITECTURE

Centralized cooperative spectrum sensing is coordinated by a common receiver [9]. All cognitive users in the network initiate local spectrum sensing independently and then forward their observations to the common receiver for a final decision. Cooperative spectrum sensing in centralized architecture is usually

performed in two successive stages: the sensing and reporting steps. Before sending the reports, some sort of authorization from the common receiver is needed.

In the sensing stage, an individual cognitive user performs spectrum sensing based on local measurements and observations. In the reporting stage, all local sensing observations are reported to a fusion receiver and the latter will make a final decision on the absence 0H or the presence 1H of the PU. The details of these stages are as follows [10]:

1) Each CR node performs local spectrum sensing

measurements independently and then makes a binary decision

2) All the cognitive users forward their binary decisions to a common receiver

3) The common receiver combines these binary decisions to infer the absence or presence of the PU in the observed frequency band according to a decision fusion rule.

If the channels between cognitive users are perfect and

decision fusion is employed at the common receiver, the detection probability dQ , the false alarm probability fQ , and the miss-detection probability mQ of cooperative spectrum sensing are then given by

( )∏=

−−=K

iidd PQ

1,11 (4)

( )∏=

−−=K

iiff PQ

1.11 (5)

where K is the number of collaborated CRs; and idP , ,

ifP , ,and imP , are the detection, false alarm, and miss-detection probabilities for the ith

A. Cooperation in Non-Fading Environment

cognitive user, respectively.

In a non-fading environment, channel case is deterministic, and false alarm probability fP is independent of instant SNR γ under the idle channel condition. The probabilities of detection, false alarm, and miss-detection are given by the following formulas [11]:

{ }[ ]11 HHPEQd γ=

{ } ( )λγλ ,21 mQHYP =>= (6)

{ }[ ]01 HHPEQ f γ=

{ } ( )( )m

mHYPΓ

Γ=>=

2,0

λλ

(7)

where [ ]⋅γE is the expectation over instant SNR γ ; ( )⋅Γ and ( )⋅⋅Γ , are complete and incomplete gamma functions,



respectively [12]; and ( )⋅⋅,mQ is the generalized Marcum Q-function [13] defined as

( ) ( )dxxIeaxQ m

xm

m

m αβαβ

α1

2/)(1

22, −

∞+−

−∫=

(8)

where ( )xIm 1− is the modified Bessel function of ( 1−m )th

B. Cooperation in Fading Environment

order.

In a fading environment, the detection probability dPis related to instant SNR γ . Therefore, by averaging Equation (6) over fading statistics, dP can written as

( ) ( )dssfQPs

md ∫= γλγ ,2

(9)

where ( )sfγ is the probability distribution function (pdf) of instant SNR γ under fading. By substituting ( )sfγ in (9), dP can be given as

( )( )

+

−×

++

=

∑

∑−

=

−+−

=

−−

2

0

212

0

12

12!1

12!

1

m

k

k

m

k

mk

d

kee

keP

γγλ

γγλ

λγ

λ

λ

(10)

From Equation (10), we can conclude that cooperative sensing can improve the overall detection performance of cognitive systems. However, the reliability of local spectrum sensing and reporting channel are crucial to realize cooperative sensing. The varying sensitivity of the cognitive nodes results in the different reliabilities of the local sensing observations of a single node, which will have a crucial effect on common receiver decision reliability.

The common detection probability of the common receiver is also related to the channel gain of the reporting channel, which is in turn related to the SNR over the link. Therefore, some nodes encounter deep fading, which may cause serious false alarm fP or miss-

detection mP . Such condition may weaken the performance of cooperative sensing when the numbers of the deep faded nodes are large. Transmitting additional data, i.e., likelihood ratio, credibility, or raw detection data, to the fusion center/common receiver and using some sophisticated fusion rules may increase the reliability of final decisions. However, the transmission of additional data requires large bandwidth for the control channel [14].

Cooperative sensing detection performance is seriously dependent on the local energy detection of a single node when all cooperative CR nodes fall under similar low levels of SNR γ . In [15], a similar situation where all CR nodes are under a low level SNR γ perform cooperative sensing is assumed. The scenario is evaluated and compared with the energy detection of a single CR node.

Their result shows that the receiver operating characteristic of cooperative sensing and single-node local sensing have similar characteristics. Therefore, as a conclusion, when CR nodes are under a severe fading environment, cooperative spectrum sensing has less benefit for improving the final decision detection performance.

To solve this problem, a cooperative spectrum sensing algorithm based on SNR-level comparison is proposed in [11]. Each CR node sends the estimated SNR to a fusion center for final decision. First, the fusion center performs SNR comparison according to the design rule. Only CR nodes with better SNR level γ are incorporated in the fusion decision. The method achieves enhanced detection performance and reduces the number of CR nodes that participate in decision fusion. Instinctively, collaborative sensing performance would degrade because of shadowing correlation when CR users are close to one another because all the nodes that are close to one another will encounter the same shadowing. The shadowing effect can be modeled as an exponential correlation function as written in [16].

( ) adedR −= (11)

where the correlation function is ( )dR , d is the distance between of two CR locations, and a is a propagation constant that depends on the wireless environment. The propagation constant is usually a = 0.1204 in urban area environments and a ≈ 0.002 in suburban environments. Correlated shadowing ( )dR in close locations (small d ) degrades the performance of collaborative detection in urban areas more than in rural areas. However, when two users are located further apart, this effect becomes less significant. Thus, when designing collaborative spectrum sensing protocols, the relative location of users should be considered.

C. Impact of Control Channel Cooperation sensing is needed by the control channel

to report information transmission. A control channel can either be implemented as a dedicated frequency channel or as an underlay channel [17]. The CR node can use a transceiver with wideband front-end tuners and filters for both regular CR reception/transmission and the underlay control channel. Furthermore, the control channel needs to be shared between active multi-user CRs using one of the multiple access techniques, i.e., CSMA/CA and CDMA. The CSMA scheme can be used as multiple-access technique for control-channel sharing among CR nodes. Moreover, the spread spectrum technique, i.e., CDMA can be used, where different spreading sequencing codes could be assigned to different groups of cognitive users.

Bandwidth limitation is considered as the main challenge and characteristic of a control channel. Control channels usually have lower power and bandwidth, such that the spectral efficiency and usage is also dependent on the control channel bandwidth. This scenario occurs when all CRs need to report their decision or local



observation, monitoring, and/or sensing information, which indicates that infinite bits are required to be reported, thus congesting the control channel [18].

To solve the bandwidth constraint in the control channel, quantization of local observations has attracted considerable research interest. However, quantization introduces additional noise and an SNR loss at the receiver [19]. Many studies have been conducted on quantization for sensed signals, but most focused on the optimal design of the quantizer [20]. Quantization of detected signals into two or three bits is the most appropriate because it can be executed without a noticeable degradation in performance [21]. Identical binary quantization, i.e., one bit quantization, performs asymptotically optimal because the number of users goes to infinity [22]. However, when the number of CRs is extremely large, the total number of sensing bits reported to the fusion center remains excessively large.

To reduce bandwidth, censoring sensors have been introduced in spectrum sensing under communication constraints where only the likelihood ratios (LR) with sufficient information are allowed to report the sensing results to a fusion center, considering perfect reporting channels [23]. Cooperative spectrum sensing with 1 bit quantization is considered is in [24]. Every cognitive user obtains an observation independently and then determines the reliability of its information. Two thresholds λ1 and λ2

≥<≤

=2

1

100

λλ

i

ii O

OD

are used to measure the reliability of the collected energy, as shown in Fig. 6. After censoring their observations, only users with reliable information are allowed to report their local binary decisions to the common receiver, whereas others will not make any decision during the reporting stage.

(12)

The false alarm probability of the method is bounded because of failed sensing, and a loss of false alarm probability is caused by the large “No Decision” region. Misdetection probability mP is degraded by the imperfect channel, and the false alarm probability fP is bounded by the reporting error probability. This condition means that spectrum sensing cannot be successfully conducted when the desired fP is smaller than the bound fP . However, the number of sensing bits is dramatically decreased, compared with the conventional method.

Figure 6. Detection method with bi-thresholds To minimize reporting channel bandwidth, a reporting

scheme that reduces the average number of reporting bits is also presented [16] by allowing only the candidate node with detection information to report its result to fusion node (FN). In this work, if Q exceeds the threshold λ , a reporting decision R is taken, and binary

decision 1 is sent to fusion node; otherwise, “no decision” R′ is taken. If the FN receives “0,” decision it auto corrects he decision to “1” because no detection case should not be reported.

Fig. 7 evaluates performance in terms of bandwidth requirement for the control channel. Average sensing channel SNR 10=γ dB is considered, and the probability of detection dP is set to 0.9. The curves in the graph show the normalized reduction of the reporting channel bandwidth based on the autocorrecting reporting method and also compares the method with the bi-threshold (two-level quantization) method (Sun et al., March 2007) and the conventional (full reporting) method.

Figure 7. The normalized average number of sensing bits k vs. fP ,

(SNR 10=γ dB, 9.0=dP )

D. Data Fusion The combination of sensing results reported by CRs or

data fusion is a key issue in sensing algorithm performance. Soft decision combinations of cognitive node observation achieve better performance. Soft combination requires the cognitive nodes to be tightly synchronized to overcome the wallSNR when the cognitive nodes are independent and identically distributed (iid). However, the wallSNR will emerge in a primary interference environment, and perfectly synchronized soft cooperation will only reject uncorrelated interference sources as well as increase the sample number.

As discussed in the previous section, hard decision combinations yield gains in terms of bandwidth limitation of the control channel. A quantization-based scheme provides good performance, but has the drawback of energy consumption at the cognitive node. The K-out-of-N fusion method [11] is a suboptimal sensing scheme that compares the total number of sensor nodes that vote for signal detection against a defined threshold K. Or-rule and AND-rule logic operations are also used for the combination of sensing results. The Or-rule fusion method avoids interference to the PU, but produces a high probability of false alarm because it declares the detection of the PU with only one cognitive user.

0 iQ

Decision 0H

Decision 1H

No Decision

1λ

2λ



E. Throughput Analysis Despite reducing the required sensing time to achieve

reliable accuracy, cooperative sensing introduces reporting and fusion time delay. Fig. 8 evaluates the sensing duration (time to sense, report, and fusion) in local sensing, direct reporting, and DCS methods. The cooperative curves show that total time is reduced when the number of nodes increases, which can be attributed to a reduction in sensing time. On the other hand, the direct reporting and DCS methods at 15 and 14 users achieve better performance compared with local node sensing. Under the described cooperation scheme, average detection time is reduced, thus implying an increase in agility. Direct reporting outperformed the DCS method by a small margin because of the lack of intermediate fusion at the FN. However, this margin is sufficiently small to be disregarded because of the parallel and distributed data fusion in the DCS system. Notably, the DCS scheme enables efficient sensing time from the idle reporting slot because it uses dynamic TDMA reporting mode, where a combination of sensing slot and idle reporting slot is used for sensing, thus reducing the original sensing slot significantly.

Figure 8. Sensing duration vs. number of nodes in Local (single user), direct reporting, and DCS method ( 1.0=fP , 9.0=fP , SNR 10=γ dB)

F. Drawbacks of Cooperative Methods Table I summarizes main characteristics of existing

cooperative methods and identifies their drawbacks.

V. CONCLUSION

Cooperative sensing is introduced as the key to reducing the probability of interference to legacy systems. Cooperative sensing requires a combination of large cooperative technologies including data fusion algorithms, data exchange protocol, and network architecture. However, cooperative gains are based on the validity/reliability of sensing, control channel, data exchange protocol, node selection, data errors, overhead management, and the cooperation protocol of the network.

ACKNOWLEDGMENT

The first author gratefully acknowledges the support of the Sudan University of Science and Technology (SUST) under the grant SUST-SRD-2012-B1024 and University Kebangsaan Malaysia (UKM), Malaysia under fundamental research grant scheme: FRGS/1/2012/SG05/UKM/02/7.

TABLE I

CETELIZED COOPERATIVE SENSING CHARACTERISTIC AND DRAWBACKS

Sensing method

Network Architecture

Problem Consideration Shortage/drawback

Energy detection

Centralized fading impact in sensing channel and hidden node problem

-Complexity in decision fusion -Impact of Reporting channel -Based on basic energy detection -Probability of detection bounded with probability of false alarm -Channel BW

Adhoc

Cluster-Based (Nuttall, 1975)

Complexity in decision fusion and Impact of Reporting channel

Cluster-Based

Design of link layer protocol and effect of node mobility

-Effects of transmission errors and nodes -connectivity on quality of detection -Protocol time and synchronization issues

Cyclost-ationary feature

Centralized

Enhance feature detection

Detection of unknown signal

Relay based

Two user/ Multi-user scenarios

PU detection Risk of interfering with PU in transmitting slot

REFERENCES [1] 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, vol. 2010, p. 2, 2010.

[2] Y. Alemseged, C. Sun, H. Tran, and H. Harada, "Spectrum Sensing for Cognitive Radio Systems: Technical Aspects and Standardization Activities, Part II," IEICE Tech. Rep, vol. 109, pp. 97-104.

[3] FCC10-174. (2010). Unlicensed Operation in the TV Broadcast Bands- Additional Spectrum for Unlicensed Devices Below 900 MHz and in the 3 GHz Band.

[4] J. N. Laneman, D. N. C. Tse, and G. W. Wornell, "Cooperative diversity in wireless networks: Efficient protocols and outage behavior," Information Theory, IEEE Transactions on, vol. 50, pp. 3062-3080, 2004.

[5] J. N. Laneman and G. W. Wornell, "Distributed space-time-coded protocols for exploiting cooperative diversity in wireless networks," Information Theory, IEEE Transactions on, vol. 49, pp. 2415-2425, 2003.

[6] G. Ganesan and L. Ye, "Cooperative spectrum sensing in cognitive radio, part II: Multiuser networks," Wireless Communications, IEEE Transactions on, vol. 6, pp. 2214-2222, 2007.

[7] G. Ganesan and Y. Li, "Agility improvement through cooperative diversity in cognitive radio," in IEEE Global Telecommunications Conference, GLOBECOM '05. , 2005, vol. 5, pp. 5 pp.-2509.

[8] K. Moessner, G. Dimitrakopoulos, P. Demestichas, J. Luo, O. Sallent, and C. Kloeck, "Dynamic radio resource allocation strategies and time scales," in SDR Forum Technical Conference 2005, Orange County, CA., 2005.

[9] F. F. Digham, M. S. Alouini, and M. K. Simon, "On the energy detection of unknown signals over fading channels,"



Communications, IEEE Transactions on, vol. 55, pp. 21-24, 2007. [10] K. Ben Letaief and W. Zhang, "Cooperative communications for

cognitive radio networks," Proceedings of the IEEE, vol. 97, pp. 878-893, 2009.

[11] A. S. Rashid, C. Bharat, and A. M. Rania, Femtocell Communications: Business Opportunities and Deployment Challenges. USA: IGI Global, 2012.

[12] A. S. Rashid, M. Rania, and K. Sabira, "Spectrum Sensing and Sharing for Cognitive Radio and Advanced Spectrum Management," ICGST International Journal on Computer Networks and Internet Research (CNIR), vol. 9, pp. 87-97, 2009.

[13] M. Gudmundson, "Correlation model for shadow fading in mobile radio systems," Electronics letters, vol. 27, pp. 2145-2146, 1991.

[14] T. Weiss, "A diversity approach for the detection of idle spectral resources in spectrum pooling systems," in Proc. of 48th International Scientific Colloquium, Ilmenau, Germany., 2003.

[15] C. W. Helstrom, "Improved multilevel quantization for detection of narrowband signals," Aerospace and Electronic Systems, IEEE Transactions on, vol. 24, pp. 142-147, 1988.

[16] A. M. Rania, S. Khatun, M. A. Borhanuddin, and A. R. Ramli, "Hard-Decision Based Reporting Method for Limited Band Control Channel in Cooperative Cognitive Radio Networks," The Open Electrical & Electronic Engineering Journal, vol. 2, pp. 62-65, 2008.

[17] D. Cabric, S. M. Mishra, and R. W. Brodersen, "Implementation issues in spectrum sensing for cognitive radios," in The Thirty-Eighth Conference of Signals, Systems and Computers, 2004, vol. 1, pp. 772-776 Vol. 1.

[18] A. S. Rashid and J. S. Stephen, TV White Space Spectrum Technologies: Regulations, Standards, and Applications. USA: CRC Press, 2012.

[19] N. Hoven, R. Tandra, and A. Sahai, "Some fundamental limits on cognitive radio," Wireless Foundations EECS, Univ. of California, Berkeley, 2005.

[20] W. Hashlamoun and P. Varshney, "Near-optimum quantization for signal detection," Communications, IEEE Transactions on, vol. 44, pp. 294-297, 1996.

[21] R. S. Blum, "Distributed detection for diversity reception of fading signals in noise," Information Theory, IEEE Transactions on, vol. 45, pp. 158-164, 1999.

[22] J. F. Chamberland and V. V. Veeravalli, "Decentralized detection in sensor networks," Signal Processing, IEEE Transactions on, vol. 51, pp. 407-416, 2003.

[23] C. Rago, P. Willett, and Y. Bar-Shalom, "Censoring sensors: A low-communication-rate scheme for distributed detection," Aerospace and Electronic Systems, IEEE Transactions on, vol. 32, pp. 554-568, 1996.

[24] C. Sun, W. Zhang, and K. B. Letaief, "Cooperative spectrum sensing for cognitive radios under bandwidth constraints," in IEEE Wireless Communications and Networking Conference, WCNC 2007., 2007, pp. 1-5.

Rania Abdelhameed (MIEEE) received her PhD degree in Communications Engineering from the University Putra Malaysia, Kuala Lumpur. She is a certified IEEE Wireless Communication Professional (IEEE WCP). She is an assistant professor with Sudan University of Science and Technology. She is the co-editor of the book

“Femtocell Communications and Technologies: Business Opportunities and Deployment Challenges”. Dr. Rania has been an IEEE member since 2001.

Rashid Saeed received his PhD majoring in Communications and Network Engineering, UPM, Malaysia. He has been a Senior Assistant Professor since 2008 in SUST, Sudan. He was senior researcher in Telekom Malaysia™, Research and Development (TMRND) and MIMOS Berhad in 2007 and 2010, respectively.

His areas of research interest include wireless broadband and WiMAX Femtocell. He was successfully awarded with 10 patents (two are US

patents) in these areas. Dr. Rashid has been an IEEE member since 2001 and is a member of IEM.

Raed Alsaqour is an assistant professor in the Computer Science Department, Faculty of Information Science and Technology, Universiti Kebangsaan Malaysia, Malaysia. He received his B.Sc. degree in Computer Science from Mu’tah University, Jordan, in 1997, M.Sc. degree in distributed system from University Putra Malaysia, Malaysia in 2003, and his PhD degree

in wireless communication system from Universiti Kebangsaan Malaysia, Malaysia, in 2008. His research interests include wireless networks, ad hoc networks, vehicular networks, routing protocols, simulation, and network performance evaluation. He also has a keen interest in computational intelligence algorithms (fuzzy logic and genetic) applications and security issues (intrusion detection and prevention) over network.

Yahia Abdallah received his PhD in Computer Engineering from Sudan University of Science and Technology, Khartoum, Sudan. He is a former ICT Minster and Associate Professor with the faculty of Computer Science and Information Technology, SUST. He was the Director of the Computer Center, SUST, as well as Co-founder and Dean of the faculty of the Computer Science and Information Technology,

SUST. He was also the Head of the Electronics Department, SUST. Dr. Yahia has numerous publications in the fields of information and communication technology.



study on energy detection-based cooperative sensing in ......gaussian noise. a denotes the...

Documents