JCS_V4_N1_RP4 Attacker Detection for Cooperative Spectrum Sensing in Cognitive Radio Network: Using Cluster Reduction Technique Sivadeep Pemma Venkata P. Venkatramana Journal on Communication Engineering and Systems 2277-5242 4 1 19 24 Box Plot Method, Dixon's Outlier Test, K-mean Clustering, Receiver Operating Characteristics In co-operative spectrum sensing system, secondary users co-operate among themselves, improving the cognitive radio spectrum sensing performance. The presence of two or more malicious user in the secondary users severely degrades the performance of the cognitive radio spectrum sensing system. In this paper, the author have studied the detection and elimination of the malicious users in a co-operative spectrum sensing system using clustering technique namely k-mean clustering and compared its performance with Dixon's outlier test and box plot test. The performance of the system with different outlier tests is shown by plotting the receiver operating characteristics curves. From this we can show that cluster removal method outperforms the Dixon's outlier test and box plot test for the case of two or three malicious users in a co-operative spectrum sensing system for cognitive radio network. November 2014 - January 2015 Copyright © 2015 i-manager publications. All rights reserved. i-manager Publications http://www.imanagerpublications.com/Article.aspx?ArticleId=3292