The Security of Physical Layer in Cognitive Radio Networks

The Security of Physical Layer in Cognitive Radio Networks Yi-cheng Yu, Liang Hu, Hong-tu Li, Yuan-mo Zhang, Fang-ming Wu, and Jian-feng Chu * Jilin University, Chang Chun 130012, China Email: yicheng_yu@126.com; {hul, chujf}@jlu.edu.cn; li_hongtu@hotmail.com; murong9r@163.com; 4464532@qq.com Abstract Cognitive Radio (CR) is a novel technology that promises to solve the lack of spectrum by allowing secondary users to use licensed band, so that they can coexist with primary users without causing interference to their communications. With the development of CR, extending to the level of network, cognitive radio networks (CRNs) emerge as the times require. Nowadays the operational aspects of CRN are being explored vigorously, and several potential security challenges for cognitive radio have gained lots of attention. In this paper, we explore the security issues on physical layer for cognitive radio networks. First, we give a brief overview of the CRNs, then we review several existing secure threats to the physical layer in CRNs and we propose a new kind of security problem. Next, we discuss the related countermeasures on how to defend against these attacks. Subsequently, we conduct an evaluation of these countermeasures, and make some future works for secure CRNs. At last, we make the conclusion. Index Terms Cognitive radio networks, physical layer, security threats, countermeasures against attacks, primary users location attack I. INTRODUCTION Due to the rapid development of wireless communication, the fixed spectrum allocation cannot meet the demand for more and more users. In view of this, the Federal Communication Commission (FCC) approves new rules to allow unlicensed users to utilize the spectrum reserved for wireless broadband services in 2010, and Cognitive Radio Networks (CRNs) have been proposed as a strong candidate to solve the problem. The concept of Cognitive Radio (CR) was first pioneered by the Joseph Mitola from software defined radio (SDR) in 1999 [1], and Mitola pointed out that cognitive radio is a special extension of software radio and more flexibility than software radio. Cognitive radio is a new research area for wireless communication [2]-[6], and lots of scholars have researched in this field in recent years [7]-[14]. In 2005, Haykins propsed a definition of cognitive radio, he hold that cognitive radio is an intelligent wireless Manuscript received August 24, 2014; revised December 25, 2013. This work was supported by the Deep exploration instrumentation and equipment development (SinoProbe-09-01-03) under Grant No.201011078. Corresponding author email: chujf@jlu.edu.cn. doi:10.12720/jcm.9.12.916-922 communication system that is aware of its surrounding environment, and uses the methodology of understanding-by-building to learn from the environment and adapt its internal states to statistical variations in the incoming RF stimuli by making corresponding changes in certain operating parameters in real time, with two primary objectives in mind: highly reliable communication, whenever and wherever needed, and efficient utilization of the radio spectrum" [15]. In 2008, Chen et al. stated that cognitive radios could take the opportunity to make communication with the spectrum holes in order to successfully facilitate useful services and applications [16]. With the development of cognitive radio, extending to the level of network, the cognitive radio network can utilize idle licensed spectrum, thereby improving the utilization of spectrum resources to meet the demand for more spectrum for wireless users. Because of the physical characteristics of CRNs where various unknown wireless devices are allowed to opportunistically access the licensed spectrum, several types of attacks on physical layer in CRNs has been attracting continuously growing attention. And we need to take security measures to combat attacks launched by malicious attackers. We then present these attacks on physical layer in CRNs and evaluate corresponding countermeasures showing their advantages and disadvantages. The rest of the paper is organized as follows. In Section 2, we present an overview of CRNs. Section 3 addresses the different types of secure threats on physical layer in CRNs. And we discuss and evaluate the countermeasures against these attacks on physical layer in Section 4. Subsequently, we make some future works for secure CRNs in Section 5. In Section 6, we make the conclusion. II. BRIEF OVERVIEW OF CRNS A. The Architecture of CRNs There are various kinds of communications in CRNs, which can be viewed as types of heterogeneous networks. In CRNs, there are three basic components: mobile station (), base station/access point () and backbone/core networks. To improve the entire network utilization effectively, three different architectures are 916

Journal of Communications Vol. 9, No. 12, December 2014 comprised of these three basic components: Infrastructure, Ad-Hoc and Mesh [16]. An infrastructure CRN has a central network entity, such as a BS in cellular networks or an AP in wireless local area networks (LANs). An can access a only in a one-hop manner. s in same cell communicate with each other through the and communications between devices in different cells are routed by the, as shown in Fig. 1. Backbone/Core Networks Fig. 1. Infrastructure architecture. In Ad-Hoc architecture, CRNs are formed by devices without infrastructure support. s can set up links between each other with different communication protocols. Each is responsible for determining its next events based on the local information that it observes. cannot predict the influence of its actions on the entire network based on its local observation, so it is important to design cooperation schemes for exchanging information from other s, as shown in Fig. 2. the wireless environment network. Next fusion centers make intelligent analysis. Then base stations make optimal decisions to make re-configuration. Last cognitive users take advantage of the spectrum resources according to the decision result, the entire process is called "cognitive cycle", which is first proposed by Joseph Mitola in 1999 [1]. As shown in Fig. 4, cognitive cycle is composed of spectrum sensing, analysis, decision-making and communication. In network-centric CRNs, cognitive users sense spectrum of the wireless environment, then upload the data to fusion centers, according to the received data, which make analysis to obtain the list of idle channels and inform base stations. When cognitive users request for allocation of available channels, base stations make decisions to assign channels to them according to certain rules. And cognitive users can communicate by the assigned channels. In distributed CRNs, cognitive users sense the spectrum of the wireless environment, then transmit the data to other users. So each of cognitive users could get the same data and reach a same list of the available spectrum. And, in accordance with certain competitive mechanism of resource allocation, available channels would be assigned to cognitive users for communication. Spectrum Hole Information Feedback Radio Environment Analysis Feedback Communication Decision Making Channel Capacity Assigned Channel Fig. 2. Ad-Hoc architecture. Fig. 4. The cognitive cycle. Mesh architecture is basically a combination of infrastructure and ad hoc.s are allowed to connect to the directly or through other neighboring s as multi-hop relay nodes, as shown in Fig. 3. III. PHYSICAL LAYER SECURE THREATS IN CRNS A. Primary User Emulation Attack In CRNs, a secondary user need to detect the primary user is active or not when he tries to occupy a specific channel. And he is allowed to use the specific band while it s not occupied by a primary user. Once the presence of a primary user is detected, the secondary user should switch channels immediately to an idle channel [17, 18]. If the secondary user detects the identical band is occupied by another secondary user, spectrum sharing mechanisms should be used in order to achieve spectrum fairly. A primary user emulation (PUE) attacker may disguise himself as a primary user by transmitting special signals in the licensed band, thus leading to other secondary users mistakenly believe primary user s existing. The secondary users who regard the attackers as primary users Fig. 3. Mesh architecture. B. The Cognitive Cycle In network-centric CRNs (such as infrastructure architecture), Cognitive radio nodes sense spectrum of 917

Journal of Communications Vol. 9, No. 12, December 2014 need to give up accessing the band. Therefore, the attack would succeed in preventing secondary users from accessing this channel. Up to now, there exist several types of PUE attacks [19], including selfish PUE attack, malicious PUE attack and some more complicated PUE attacks. In a selfish PUE attack, two attackers establish an appropriative link between them simultaneously to increase their share of the spectrum resources. In a malicious PUE attack, the attacker s goal is to prevent the transmission of the secondary users without using the vacant channel. And in some more complicated PUE attacks, malicious node is capable of attacking the network only when the primary user is off, so that they can save energy to do more effective attacks. Fig. 5 shows the PUE attack in network-centric CRNs. Malicious Malicious user user CRNs, sensing terminals learn the history, feedback and adjust radio parameters according to current environment in the process of spectrum sensing, analysis, decisionmaking and communication. In a learning attack (LA) [21], learning radio learns false sensory input provided by attacks. And these wrong ideas about the transmission schemes would be studied all the way. This would exert long-term effects on subsequent operations, and it can t guarantee the best fit with the actual network environment. Usually, learning attack would be launched combining with other types of attacks such as PUE attacks and OFA attacks. D. Jamming Because of the available channel s open, malicious nodes may randomly attack some available channels by jamming. There by, interrupting the communication between CRs. The malicious node may continuously transmit high-power signals on multiple channels. At this time, the user is assigned to an idle channel, but the normal communication can not be performed. When Multiple CRs are under jamming attacks in the same channel, they would feedback the experience of poor quality to BS. This would reduce the allocation for this channel, so that a malicious user can communicate with this channel [22]. Terminal Terminal Final spectrum sensing result Terminal Terminal Fusion Fusion center center Primary Primary user user Terminal Terminal Fig. 5. Primary user emulation attack in network-centric CRNs. B. Objective Function Attack Cognitive radio is flexible, and is capable of sensing the external environment, learning from the history, and making intelligent decisions to adapt to the changing environment [20], The cognitive engine in the adaptive cognitive radio has the ability to tune many radio parameters to meet specific requirements such as high transmission data rate, low delay, high security level and low power consumption. Such radio parameters include bandwidth, power, modulation type, coding rate, MAC protocol, routing schemes, encryption mechanisms, and frame size [21]. These parameters are calculated by solving one or more objective functions, however, some objective function are directly related to the inputs of users in the channel. When cognitive engine is running to calculate the radio parameters appropriate to the current environment, the attacker can make the results tailored and biased through same way. The attack is called objective function attack (OFA) [21] and the process is shown in Fig. 6. Attacker A Attacker A Attacker A fuse Ultimate object optimize... Attacker B (1) Method A Sub-goal object Sub-goal Sub-goal object PU PU Parameters setting (2) Method B Sub-goal object Attacker Fig. 7. Primary users location attack. Fig. 6. Objective function attack. E. Eavesdropping In an eavesdropping attack, a malicious node would listen to the transmission of the legitimate users, fusion C. Learning Attack Cognitive radio is smart, intelligent and cognitive. In 918

center and base station for obtaining some information about available channel. And the attacker may calculate channels which would be switched according to the algorithm of channel selection. Eavesdropping itself does not have a negative effect on cognitive radio network. However, as an assistence to achieve the purpose of attackers, it s the basis of other attacks. F. Primary Users Location Attack In addition to the secure threats above have been proposed, we propose a new kind of attack which can obtain the location of primary users, thereby launching a direct physical attack on the equipment. In CRNs, each user can detect the signal emitted by primary user, an attacker can calculate the range of the distance between the primary user and itself according to the strength of the signal. When multiple attackers estimate the position of the primary user with this method, they can get a crossover region to narrow down the area where the primary user is located in, as shown in Method A in Fig. 7. Meanwhile, the attackers can narrow the scope further by their mobility, and ultimately get the primary user's location as shown in Method B in Fig. 7. According to the location of primary user, the attacker can find the primary user and launch a physical attack on it directly, which would make the primary user disable. IV. COUNTERMEASURES AGAINST ATTACKS ABOVE In this section we will discuss the schemes against secure threats mentioned above, we first list the recent countermeasures and then evaluate them. A. Defending Against PUE Attack To defend against PUE attacks, the identity of the transmitting source needs to be identified, and many solutions have been proposed to determine the identity of the signal source [23]-[25], [27]-[30]. Chen et al. successively proposed several solutions to defend against PUE attack: Distance Ratio Test (DRT) [23], Distance Difference Test (DDT) [23] and Localization-based Defense (LocDef) [24]. In these schemes, secondary users determine whether the signal is transmitted by primary users by estimating the position of the signal source and observing the sign's characteristics. They employ RSS-based localization that exploits the relationship between signal strength and a transmitter location to obtain the location of the signal. However, these solutions could only be applied to the CRNs in which primary user's location is fixed, but the networks where mobile primary users cannot be applied. Jin et al. presented an analysis using Fenton s approximation and Wald s sequential probability ratio test (WSPRT) to detect PUE attack [25]. They considered a fading wireless environment and derived expressions for the probability of successful PUE attack employing Fenton s approximation. Then they made use of Markov inequality to provide a lower bound on the probability of successful PUE attack. Finally, WSPRT was exploited to detect PUE attack. However, this method has strict requirements on the primary user. A primary user should be located at some distance from all the users, and position fixing. In 2012, Yuan et al. proposed a defense strategy against the PUE attack in CRNs using belief propagation [26]. In their scheme, each secondary user calculates the local function and the compatibility function, computes the messages, exchanges messages with the neighboring users, and calculates the beliefs until convergence. Then, the PUE attacker would be detected, and all the secondary users in the network would be notified in a broadcast way about the characteristics of the attacker s signal. Simulation results show that this approach converges quickly, and is effective to detect the PUE attack. B. Defending Against Objective Function Attack The research on objective function attack is not mature, no good solution has been suggested to defend against it. In [31], Olga León et al. suggested defining threshold values for every updatable radio parameter and only when the parameters meet the thresholds can communication start. However, there is no mature scheme with this method. In 2011, Pei et al. proposed an appropriate proposal called MOP (Multi-objective programming model) [32]. After attackers obtain all the parameters, secondary users detect them and compare them with fitness value with MOP to decide whether attackers exist. If so, then secondary user readjusts the tampered parameters to optimal settings. This scheme is able to detect the specific objective function which is tampered, and make reasonable regulation. Even if attackers obtain the parameters, they cannot tamper the parameters. So this scheme can effectively resist OFA if MOP is security. C. Defending Against Learning Attack Some suggestions have been made to defend against learning attacks [21]. First, the learning results must always be under constant reevaluation. A feedback loop should constantly be updating learned relationships between cognitive radio inputs and outputs. Second, there should be a truly controlled environment during the learning phases, which means no adversarial signals are present during the learning phase. Third, the action which breaks to some basic theoretic results should not be exploited. Fourth, cognitive radios had better make use of group learning instead of individual learning. Therefore the attacker cannot launch a learning attack so easily. D. Defending Against Jamming To mitigate the jamming attacks, several solutions have been proposed [33]-[39]. Secondary users need to detect that a jamming attack really exists in order to counter jamming attacks. In [33] the author suggested collecting enough data of the noise to detect jamming attacks. So secondary users can differentiate the 919

Journal of Communications Vol. 9, No. 12, December 2014 interference of an attacker from normal noise when attackers try to jam secondary users and transmit large power interference. Spread spectrum (SS) [34] and frequency hopping (FH) [35] techniques also turn out to be effective against jamming attacks. SS makes the signal more robust to interference by spreading it over a large frequency band. In the scheme using FH, Whenever the secondary users find the jamming attack, they would switch to other channels that are not jammed with their high switching ability. FH is capable of reducing the probability that the frequency involved in the current communication is targeted by the jammer. convex max-min problem to maximize secrecy capacity without interfering with primary users. The maximum achievable secrecy rate can be obtained by optimizing the transmit covariance matrix in the case of Gaussian input. Algorithms were proposed to compute the maximum achievable secrecy rate for the case of single-antenna eavesdroppers, and bounds on the achievable secrecy rate were obtained for general cases with multi-antenna secrecy and eavesdropper receivers. Secrecy rate can be improved in the scheme. F. Defending Against Primary Users Location Attack Since attackers estimate the range of the distance between primary user and themselves according to the strength of the signal, for defending against this kind of attack, the signal s strength and the distance between them should not have a linear relationship. To our point of view, primary users can resist the attacks by changing the intensity of signals irregularly. E. Defending Against Eavesdropping There is still no perfect solution to effectively resist learning attack. Zhang et al. proposed an idea employing power control algorithms to increase the rate among the legitimate users while decreasing the rate to the eavesdroppers [40]. In this scheme, secondary users exploit multiple input multiple output (MIMO) transmission, primary users employ a single antenna, and eavesdroppers can use either multiple antennas or a single antenna. They studied the achievable rates of the MIMO secrecy rate between secondary users, and formed a non- G. The Evaluation of the Countermeasures In Table I, we present the evaluation of the attacks countermeasures of the physical layer. TABLE I: PHYSICAL LAYER ATTACKS, COUNTERMEASURES AND EVALUATIONS IN CRNS Attack PUE Countermeasure Evaluation Distance Ratio Test (DRT) [23] Secondary users can detect PUE attack in the process of spectrum sensing using this method, but the solution could only be applied to the CRNs in which primary user's location is fixed Distance Difference Test (DDT) [23] Be identical to DRT Localization-based Defense (LocDef) [24] Be identical to DRT Use Fenton s approximation and Wald s sequential probability ratio test (WSPRT) to detect PUE attack [25] Secondary user can detect PUE attack alone, but primary users should be located at some distance from all the users, and position is fixed The solution converges quickly, and is effective to detect the PUE attacker It is important to set the thresholds in this scheme, and the fixed thresholds may be unreasonable If the MOP is secure, this scheme can effectively resist OFA The suggestion can defend the attack to a certain extent It is hard to define the appropriate amount of data should be exploited to build the model SS makes the signal more robust to interference Use belief propagation [26] OFA Define threshold values for every updatable radio parameter and only the parameters meet the thresholds can communication start [31] MOP (Multi-objective programming model) [32] LA The learning results must always be reevaluated over time [21] Collect enough data of the noise to detect jamming attacks [33] Jamming Spread spectrum (SS) [34] Frequency hopping (FH) [35] Frequency hopping is good for cognitive radios Eavesdropping Power control [40] Secrecy rate can be improved Primary Users Location Attack Changing the intensity of signals irregularly Location information can be protected The current schemes are designed for resisting a certain attack on physical layer in CRNs. Even if the scheme can effectively resist a certain attack, it cannot withstand other attacks on physical layer. The next step, it is vital to establish a sound scheme to resist all of attacks. V. THE FUTURE WORKS FOR THE SECURE CRNS In this section, we focus on the future works of the security of the physical layer in CRNs to make cognitive radio networks safer and work more effectively. 920

In network-centric CRNs, base station gets the current environmental information via secondary users sensing. In one case, malicious nodes may tamper the data submitted to base station. In the other case, some malicious secondary users may transmit some wrong data to base station. Adding the role of primary user into the cognitive cycle makes it more reasonable. For example, when the primary user needs to exploit a particular channel, it can notify the base station in advance. Hence, when secondary users attempt to cheat the base station into believing that the channel is idle, it would not be succeed. It is important to trace to the transmitter in CRNs. When receiving a deceived data, the recipient should be capable of finding the transmitter. However, the signature mechanism is not applied to CRNs, so it is vital to solve this problem. For making a better judgment, base stations make decisions referring to the history in network-centric CRNs. However, once the wrong information has been studied, which would always affect the final decision. In my opinion, base station should evaluate the history regularly. VI. CONCLUSIONS Cognitive Radio (CR) technology is one of the strong candidate technologies to alleviate spectrum shortage problem in wireless communications. The development of cognitive radio leads to the research of cognitive radio networks. This research work focuses on the physical layer security issues of Cognitive Radio Networks (CRNs). Not only the most important secure threats on physical layer have been addressed, but also a new kind of attack has been proposed in this paper. Then we discuss the related countermeasures on how to defend against these attacks. Finally, we conduct an evaluation of these countermeasures, and make some future works for the secure CRNs. ACKNOWLEDGMENT This work was supported in part from Deep exploration instrumentation and equipment development (SinoProbe-09-01-03). REFERENCES [1] J. Mitola and G. Q. Maguire Jr, Cognitive radio: Making software radios more personal, Personal Communications. IEEE, vol. 6, no. 4, pp. 13 18, 1999. [2] S. S. Ivrigh, S. Siavash, and S. M. S. Sadough, Spectrum sensing for cognitive radio networks based on blind source separation, KSII Transactions on Internet & Information Systems, vol. 4, pp. 613 631, 2013. [3] G. P. Joshi, S. Y. Nam, and S. W. Kim, Cognitive radio wireless sensor networks: applications, challenges and research trends, Sensors, vol. 9, pp. 11196-11228, 2013. [4] V. Gardellin, S. K. Das, and L. Lenzini, Coordination problem in cognitive wireless mesh networks, Pervasive and Mobile Computing, vol. 1, pp. 18-34, 2013. [5] B. Luca, O. Marina, R. Mirco, et al., Comparison among cognitive radio architectures for spectrum sensing, EURASIP Journal on Wireless Communications and Networking, 2011. [6] S. Chang, K. Nagothu, B. Kelley, et al., A Beamforming Approach to Smart Grid Systems Based on Cloud Cognitive Radio, 2014. [7] J. Feng, G. Lu, and X. Min, Social incentives for cooperative spectrum sensing in distributed cognitive radio networks, KSII Transactions on Internet and Information Systems, vol. 2, pp. 355-370, 2014. [8] K. Arshad, R. MacKenzie, U. Celentano, et al., Resource management for QoS support in cognitive radio networks, Communications Magazine. IEEE, vol. 3, pp. 114-120, 2014. [9] S. M. Dudley, et al., Practical issues for spectrum management with cognitive radios, Proceedings of the IEEE, vol. 102, no. 3, pp. 242-264, 2014. [10] J. Huang, H. Zhou, Y. Chen, et al., Distributed and centralized schemes for channel sensing order setting in multi-user cognitive radio networks, Wireless Personal Communications, vol. 2, pp. 1-20, 2013. [11] G. Ding, Q. Wu, Y. Zou, et al., Joint spectrum sensing and transmit power adaptation in interference-aware cognitive radio networks, Transactions on Emerging Telecommunications Technologies, 2012. [12] Z. Feng, Z. Wei, Q. Zhang, et al., Cognitive information metrics for cognitive wireless networks, Chinese Science Bulletin, vol. 17, pp. 2057-2064, 2014. [13] P. S. M. Tripathi, A. Chandra, and R. Prasad, Deployment of cognitive radio in India, Wireless Personal Communications, vol. 3, pp. 523-533, 2014. [14] D. Cabric, S. M. Mishra, and R. W. Brodersen, Implementation issues in spectrum sensing for cognitive radios, in Proc. Thirty- Eighth Asilomar Conference on Signals, Systems and Computers, vol. 1, 2004. [15] S. Haykin, Cognitive radio: Brain-empowered wireless communications, Selected Areas in Communications, vol. 2, pp. 201-220, 2005. [16] K. C. Chen, Y. J. Peng, and N. Prasad, et al., Cognitive radio network architecture: Part I--general structure, in Proc. 2nd International Conference on Ubiquitous Information Management and Communication, 2008, pp. 114-119. [17] Q. Liu, Z. Zhou, C. Yang, et al., The coverage analysis of cognitive radio network, in Proc. 4th International Conference on Wireless Communications, Networking and Mobile Computing, 2008, pp. 1-4. [18] A. N. Mody, R. Reddy, T. Kiernan, et al., Security in cognitive radio networks: An example using the commercial IEEE 802.22 standard, in Proc. Military Communications Conference, IEEE, 2009, pp. 1-7. [19] W. El-Hajj, H. Safa, and M. Guizani, Survey of security issues in cognitive radio networks, Journal of Internet Technology, vol. 2, pp. 181-198, 2011. [20] Q. Mahmoud, Cognitive networks: Towards self-aware networks, Wiley E-Book, New York, 2007. [21] T. C. Clancy and N. Goergen, Security in cognitive radio networks: Threats and mitigation, in Proc. 3rd International Conference on Cognitive Radio Oriented Wireless Networks and Communications, IEEE, 2008. [22] A. Sampath, H. Dai, H. Zheng, et al., Multi-channel jamming attacks using cognitive radios, in Proc. 16th International Conference on Computer Communications and Network, IEEE, 2007. [23] R. Chen and J. M. Park, Ensuring trustworthy spectrum sensing in cognitive radio networks, in Proc. 1st IEEE Workshop on Networking Technologies for Software Defined Radio Networks, IEEE, 2006, pp. 110 119. [24] R. Chen, J. M. Park, and J. H. Reed, Defense against primary user emulation attacks in cognitive radio networks, Selected Areas in Communications, vol. 1, pp. 25-37, 2008. [25] Z. Jin, S. Anand, and K. P. Subbalakshmi, Detecting primary user emulation attacks in dynamic spectrum access networks, in Proc. IEEE International Conference on Communications, 2009, pp. 1 5. [26] Z. Yuan, D. Niyato, H. Li, et al., Defeating primary user emulation attacks using belief propagation in cognitive radio networks, Selected Areas in Communications, vol. 10, pp. 1850-1860, 2012. 921

[27] S. Chen, K. Zeng, and P. Mohapatra, Hearing is believing: Detecting mobile primary user emulation attack in white space, in Proc. INFOCOM, IEEE, 2011, pp. 36-40. [28] D. Hao and K. Sakurai, A differential game approach to mitigating primary user emulation attacks in cognitive radio networks, in Proc. 26th International Conference on Advanced Information Networking and Applications, IEEE, 2012, pp. 495-502. [29] D. Pu, Y. Shi, A. V. Ilyashenko, et al., Detecting primary user emulation attack in cognitive radio networks, in Proc. Global Telecommunications Conference, 2011, pp. 1-5. [30] C. Xin and M. Song, Detection of PUE attacks in cognitive radio networks based on signal activity pattern, Mobile Computing, vol. 5, pp. 1022-1034, 2014. [31] O. León, J. Hernández-Serrano, and M. Soriano, Securing cognitive radio networks, International Journal of Communication Systems, vol. 5, pp. 633-652, 2010. [32] Q. Pei, H. Li, J. Ma, et al., Defense against objective function attacks in cognitive radio networks, Chinese Journal of Electronics, vol. 1, pp. 138-142, 2011. [33] W. Xu, T. Wood, and W. Trappe, et al., Channel surfng and spatial retreats: Defenses against wireless denial of service, in Proc. 3rd ACM Workshop on Wireless Security, Philadelphia, PA, January 2004, pp. 80-89. [34] A. Attar, H. Tang, A. V. Vasilakos, et al., A survey of security challenges in cognitive radio networks: Solutions and future research directions, Proceedings of the IEEE, vol. 100, no. 12, pp. 3172-3186, 2012. [35] S. Liu, L. Lazos, and M. Krunz, Thwarting inside jamming attacks on wireless broadcast communications, in Proc. Fourth ACM Conference on Wireless Network Security. ACM, 2011, pp. 29-40. [36] W. Wang, S. Bhattacharjee, M. Chatterjee, et al., Collaborative jamming and collaborative defense in cognitive radio networks, Pervasive and Mobile Computing, vol. 4, pp. 572-587, 2013. [37] M. Camilo, et al. Anti-jamming defense mechanism in cognitive radios networks, in Proc. Military Communications Conference, 2012, pp. 1-6. [38] R. Di Pietro and G. Oligeri, Jamming mitigation in cognitive radio networks, Network, IEEE, vol. 3, 2013. [39] L. Zhang, Q. Pei, and H. Li, Anti-jamming scheme based on zero pre-shared secret in cognitive radio network, in Proc. Eighth International Conference on Computational Intelligence and Security, IEEE, 2012, pp. 670-673. [40] L. Zhang, R. Zhang, Y. C. Liang, et al., On the relationship between the multi-antenna secrecy communications and cognitive radio communications, IEEE Trans on Communications, pp. 1877-1886, 2010. Yi-Cheng Yu was born in Fujian, China in 1990. He received the B.S. degree from the Jilin University, Changchun in 2012, and he is currently working toward the M.S. degree at Jilin University. His main research interest includes computer networks and information security. Liang Hu had his BE ng on Computer Systems Organization in 1993 and his PhD on Computer Software and Theory in 1999. He is currently Professor and PhD supervisor of Jilin University, China. His main research interest includes network security and distributed computing. As a person in charge or a principal participant, Dr Liang Hu has finished more than 20 national, provincial and ministerial level research projects of China. Hong-tu Li was born in Siping of Jilin, China on Mar. 17 1984. Now he is the teacher of the Jilin University, Changchun, China. He received the Ph.D. degree in computer structure from Jilin University in 2012. His current research interests focus on network security and cryptology. Yuan-mo Zhang was born in Jilin, China in 1989. She received the B.S. degree from the Jilin University, Changchun in 2012, and she is currently working toward the M.S. degree at Jilin University. Her research interest includes computer networks and information security. Fang-ming Wu received his B.S. degree from the PLA Information Engineering University in 2007. He is currently pursuing in the College of Computer Science and Technology, Jilin University. His research interest is the communication of WSN, computer networks and information security. Jian-feng Chu was born in 1978, Ph.D., Now he is the teacher of the College of Computer Science and Technology, Jilin University, Changchun, China. He received the Ph.D. degree in computer structure from Jilin University in 2009. His current research interests focus on information security and cryptology. 922