Research On Security And Privacy In Cognitive Radio Networks

For the rapid development of wireless communication technology, parts ofspectrum bands are too crowded to satisfy the increasing demands of wireless users. Butthe survey shows that most authorized spectrum bands are idle in most of the time. Tomake full use of spectrum bands to meet more wireless users, cognitive radio networksappeared. Cognitive radio networks can provide spectrum access opportunities forunauthorized users (cognitive users) without interfering authorized users (primary users).Thus it has a bright future in wireless networks.How to get the spectrum hole information and allocate it is a critical problem incognitive radio networks. For the variability and unstability of primary users in wirelessenvironment, the accurate and reliable spectrum sensing becomes the prerequisite foroperation in cognitive radio networks. But there exists malicious user who can falsifylocal sensing data to destroy authenticity of the spectrum sensing data, and finally toimpact normal operation of cognitive radio networks. For the dynamic and openness ofspectrum holes, attackers can destruct spectrum sensing and accessing to result in DoSattack. It would prevent cognitive users to access spectrum holes, and result in the wasteof spectrum resources. Besides, how to allocate available channels without interferenceamong cognitive users and maximum the privacy protection is a big challenge forcognitive radio networks. In this paper, we research on the security and privacy incognitive radio networks. The author’s major contributions are outlined as follows:1) For the collusion attack in clustering cognitive radio networks, we propose areputation based sub-spectrum sensing scheme. Through binding of spectrum bandsand dividing cognitive users into groups, the proposed scheme can do one to onemapping spectrum sensing and make final decision with consideration of reputation.It could detect malicious users timely and defense against collusion attack toensure the reliability of spectrum sensing.2) For the spectrum sensing data falsification attack in cognitive radio networkswithout infrastructure, we propose a neighbor detection based consensus spectrumsensing scheme. With the proposed scheme, cognitive users can make samedecision of spectrum sensing data in the local network through interaction withneighbors. During the convergence process, we enhance the accuracy of sensingdata by updating the trust neighbor sets for each cognitive user. We detect theneighbors sensing data in each updating round with neighbor detection mechanism to discover the destroyed sensing data and make the final decision more reliable.3) For tracking user attack during spectrum allocation in cognitive radio networks, wepropose an ordered set based channel selection scheme. During channel applicationprocess, in consideration of information leaking, we use the corresponding numberof each channel in user’s prefer sets to instead of the channel itself to complete thechannel application. It can confuse attackers and conceal the information ofselected channels effectively. We also design a synchronous update mechanism ofprefer sets to change the number corresponded to each channel dynamically toincrease the difficulty of getting spectrum utilization for attackers. The proposedscheme can hide the channel selection information to prevent attackers trackingcognitive users effectively and ensure accessing for cognitive users4) For the location privacy leaking during spectrum sensing in data-base drivencognitive radio networks, we propose a location privacy protection channelselection scheme with self-coexistence mechanism. First, we adopt self-coexistencemechanism to allocate channels for each base station to reach the purpose ofspectrum sharing without interfering among cognitive uers. We use Hash matchingto identify cognitive users without leaking cognitive users’ identities. Besides, weprevent spectrum utilization information leaking by base stations’ registration.Therefore, even attackers get the register information from data base, they cann’tconnect the information with the corresponding cognitive user. Therefore, theproposed location privacy protection scheme can protect location privacy ofcognitive users and make full use of available spectrum.