Nullifying Malicious Users for Cooperative Spectrum Sensing in Cognitive Radio Networks Using Outlier Detection Methods

摘要

A number of cooperative spectrum sensing techniques have been purposed in cognitive radio networks. However, collaboration between multiple cognitive radio (CR) users also raises a number of security issues. It has been shown that the cooperative gain can be severely affected by malfunctioning or malicious CR users in cooperative sensing. One of them is spectrum sensing data falsification (SSDF) attack, where malicious users transmit false information instead of real detection results and thereby affecting the final decision. In this paper, we study the detection and suppressing the malicious users using different outlier detection methods based on Grubb's test, Boxplot method and Dixon's test. We have compared their performance through simulation and receiver operating characteristics (ROC) curve shows that Boxplot method outperforms both Grubb's and Dixon's test for the case where multiple malicious users are present.