Detecting Primary User Emulation Attacks In Cognitive Radio Networks

This thesis first covers an introduction of cognitive radio, dynamic spectrum access and spectrum sensing, as well as the potential security threats and their impacts to cognitive radio networks. Then we focus on primary user emulation (PUE) attacks and investigate existing techniques for detecting this threat such as energy detection, localization, cyclostationary feature calculation and artificial neural network. To counter the PUE attack, an improved energy detection method with multiple thresholds is proposed. Then we propose two new schemes that combine the improved energy detection method with TODA localization and cyclostationary feature detection respectively to identify PUE attackers in different scenarios in cognitive radio networks.