Performance Analysis of Voting Rule in the Presence of SSDF Attack in Interweave Cognitive Radio Networks

Abstract

Cognitive radio (CR) is widely recognized as a viable solution for spectrum underutilization. Cooperative spectrum sensing (CSS) is a core aspect of CR that improves spectrum sensing performance through spatial diversity. Howeverr, CR provides malicious users (MUs) to launch spectrum sensing data falsification (SSDF) attacks, thus degrading the performance of CSS. And the reporting channel from secondary users (SUs) to the fusion center (FC) consumes significant communication overhead in large CR networks (CRNs), thus reducing CSS efficiency. In front of this issue, we take voting rule in the presence of SSDF attack into consideration to enhance the security and efficiency of CSS. To this end, we make an in-depth analysis on conventional voting rule (CVR) and sequential voting rule (SVR) in an interweave CRN and propose a sequential 0/1 (S0/1) rule to reduce the number of samples. Furthermore, the closed-form performance expressions of S0/1 are also theoretically derived in the absence or presence of SSDF attack. Finally, in contrast to CVR and SVR, numerical simulation results show that the proposed S0/1 is able to maintain the same CSS performance and better CSS efficiency in the presence of SSDF attack.