A Cooperative Spectrum Sensing Method Based on Multiple Reference Information Sources and Attack Recognition to Defend Against SSDF

Cooperative spectrum sensing based on reputation algorithm with global decision-making as reference information. In the large-scale Spectrum Sensing Data Falsification (SSDF) attack scenario (the proportion of malicious users (MUs) is greater than $50 \%$), due to the loss of the reliability of the reference information, the reputation algorithm’s ability to resist SSDF attacks is completely lost. In order to ensure that cooperative spectrum sensing has better anti-attack ability under large-scale SSDF attacks, an anti-SSDF attack cooperative spectrum sensing method based on multi-reference information sources and attack recognition (MRISAR-CSS) is proposed. Firstly, according to the historical information of the secondary user (SU) recorded by the data fusion center (FC), the behavior similarity and reputation value are calculated, and all SUs are screened by different reputation values to obtain multiple SU sets. Then, the SSDF attack strategy and attack probability are identified by historical information and behavior similarity, and the reliability of multiple SU sets obtained by reputation value screening is evaluated to obtain the set with the highest reliability in this SSDF attack scenario. Finally, according to the estimated relevant parameters, the most suitable data fusion method under this condition is determined to make FC get the global decision. The simulation results show that the proposed method has good perceptual accuracy in the face of large-scale SSDF attacks with different attack strategies and attack probabilities and has better antiattack ability than traditional methods.