Defending Against Byzantine Attacks in CRNs: PCA-Based Malicious User Detection and Weighted Cooperative Spectrum Sensing

Cognitive radio (CR) technology is a viable solution for assisting secondary users to share the licensed radio spectrum of primary users. Cooperative spectrum sensing (CSS) enhances the accuracy of spectrum sensing in a CR network. However, the effectiveness of CSS can be compromised by malicious users (MUs) who intentionally send false sensing information to the fusion center. This letter focuses on enhancing the CSS performance and detecting the MUs. We propose a machine learning technique to identify and classify MUs in a CR network using the Principal Component Analysis algorithm. The performance of the proposed algorithm in detecting MUs and enhancing CSS performance is validated through simulation experiments.