Malicious node(s) identification in Ad-Hoc cognitive radio networks (CRN’s)

Abstract

Cognitive Radio Networks (CRNs) are emerging technology to increase the efficiency of wireless spectrum. In CRN, two types of users are utilizing the radio frequency (RF) spectrum: primary user (PU) who have priority of using spectrum and secondary user (SU) who uses spectrum opportunistically.Cognitive behavior of these network introduced new security challenges with traditional security issues. Spectrum sensing plays vital role in performance of CRN. Attacks during spectrum sensing phase like Spectrum Sensing Data Falsification (SSDF) attacks cause severe damage that can threaten the existence of the CRN. In distributed cognitive ad hoc radio network, these attacks are more difficult to detect due to mobility and lack of central body.The presented work proposed a framework and reputation management system for detection of malicious nodes that launched the SSDF attacks. The three variants of SSDF attack are modeled and performance of reputation system is analyzed using different simulation parameter settings. The simulation results show 99.4 % spectrum decision accuracy. The malicious node detection accuracy converges to 100 %.