Improving detection of primary user in a cognitive radio network with malicious users

Abstract

A cooperative spectrum sensing (CSS) scheme with multiple malicious secondary users (MSUs) is considered. These MSUs carry out spectrum sensing data falsification (SSDF) attacks. An MSU suppression scheme is considered which consists of an improved energy detector (IED) followed by a statistical algorithm implemented at the fusion center. Three statistical algorithms are considered which are: the Tietjen-Moore (TM) test, the Extreme Studentized Deviate (ESD) algorithm and the Pierce criterion. The simulation results of an SSDF attack show that the performance of an IED followed by either TM or ESD test is better than that of the standard energy detector followed by the same algorithm, respectively.