Performance Comparison of Hard Decision Schemes in the Presence of Malicious Users

Abstract

Cognitive Radio Network (CRN) is an intelligent technique to access the radio spectrum. In CRN the Secondary Users (SUs) sense the spectrum hole of the Primary User (PU) band opportunistically. Spectrum sensing is the major step in CRN to observe the Primary User (PU) signal. For individual user, sensing ability is affected by fading and shadowing. Cooperative Spectrum Sensing (CSS) provide solution to the problem of fading and shadowing where multiple SUs cooperates with each other to catch the absence and presence of PU signal. However, it is vulnerable to a variety of malicious User (MU) attacks. These MUs are unfriendly to the sensing results that could be achieved by CSS. In the proposed work performance of hard combination schemes are tested under varying Signal to Noise Ratios (SNRs). The performance of the hard combination schemes such as majority voting, logical-AND and logical-OR are tested in the presence of Lazy Users (LUs) and Opposite Malicious Users (OMU) at different SNR levels. Results indicate that hard combination scheme such as majority voting outperforms all other schemes, such as logical-OR and logical-AND.