A new reputation algorithm based on sliding window with punishment mechanism

Spectrum sensing data falsification (SSDF) attack is a serious threat to collaborative spectrum sensing (CSS). In this paper, in order to detect malicious users (MUs) more quickly, and to alleviate the performance deterioration in CSS caused by SSDF attack, a new reputation algorithm based on sliding window with punishment mechanism is proposed. The proposed algorithm integrates the distribution information of subjective false-report time-slots over a sliding window into appraisal of current reputation. Through the design of indicator function sliding window and a penalty factor, we develop a new reputation calculation method to punish the user that frequently performs false report, and provides a fast WSPRT (FWSPRT) algorithm by adopting the new reputation algorithm. Simulation results show that the proposed reputation algorithm can recognize MU more quickly and accurately, and FWSPRT provides a better performance than WSPRT.