An Empirical Estimation of CSS Cognitive Radio Network Performance under Spectrum Sensing Data Falsification Attack

Cooperative spectrum sensing (CSS) significantly improves the performance of spectrum sensing process in cognitive radio networks (CRNs). Individual spectrum sensing by a cognitive radio (CR) is often inaccurate as the channel often experiences fading and shadowing effects. CSS has been shown to have many advantages in terms of spectrum use and robustness. Despite these facts, a CSS scheme also vulnerable to many security attacks from malicious users (MUs). In order to get unfair usage of spectrum band, MUs can generate false spectrum sensing reports to disturb the good secondary users (SUs) decision about presence of primary user (PU). In this paper, we consider the spectrum sensing data falsification attack (SSDF) in CSS and propose the protocol to identify and eliminate the attacker or malicious user (MU) to improve the network performance. In SSDF attack, MUs send the false spectrum sensing results to fusion center (FC) with the intension that it should make wrong decision about spectrum sensing. In this scenario, FC acts as a data collector to fuse the reports sent by SUs.