Anomaly user effect on detection performance of eigenvalue and volume based spectrum sensing

In wireless communication networks, the physical layer security is of prime concern which is mainly effected due to anomaly/malicious users. Cognitive radio network (CRN) is one of the next generation wireless networks, recently emerged as a technology that can use unoccupied licensed frequency bands temporarily without taking permission from spectrum regulatory bodies. Owing to inherent nature of unlicensed access, CRNs are more likely to anomaly user effects that degrades the Quality of service (QoS) of both licensed and cognitive user communications. These effects are generally taken place during spectrum sensing. To eliminate single anomaly effects and improve accuracy of sensing decision by applying spatial diversity concept, cooperative sensing techniques have been developed. However, the detection probability of cooperative sensing degrades owing to the presence of multiple anomaly user in the network. In particular, due to spectrum sensing data falsification (SSDF) attacks. Hence, in this work, three types of basic SSDF anomaly in CRNs are investigated and presented their impact on detection probability by selecting two prominent signal detection methods namely Volume detection (VD) and Max-Min eigenvalue detection (MME). Simulation results reveal that there is a significant performance degradation due to these anomaly that has to be addressed to improve the detection performance.