The Analysis of Influence on Sensing Performance from the Malicious user for SSDF Attacks in Cognitive Radio

Cognitive radio (CR) can improve the utilization of the spectrum by making use of licensed spectrum. However, the security aspects of cognitive radio networks have garnered little attention. In this paper, we discuss a threat to cognitive radio networks, which we call the spectrum sensing data falsification (SSDF) attack. Specifically, malicious users can always send presence and absence of information for the authorized user, or always report the opposite sensing results. Collaborative sensing performance was deduced under different attack scenarios. Simulation results show that compromise collaborative sensing performance can be obtained when a malicious user always sends "1" with the AND rule, and always sends "0" with the OR rule; In addition, for a malicious user always report the opposite sensing results, q-out-of-N rule was adopted for the cooperative sensing. When the number of malicious users is less, q values should be smaller. On the other hand, from the perspective of make full use of spectrum resources, q values should be larger. Key-Words: Cognitive radio, Spectrum Sensing, SSDF attacks, Attack scenario, malicious user, fusion rule