A complete algorithm to diagnose and alleviate the effects of physical layer attacks

In this paper we present a method to diagnose and mitigate against primary user emulation attacks (PUEA) in cognitive radio networks. We develop a hybrid algorithm that uses a combination of compressed sensing and belief propagation to identify and combat PUEAs. We propose to use compressive sensing at the fusion centre to localise a primary user, then distribute the primary user location to secondary users in order to establish theoretical data for comparison and then use a variant of belief propagation at each secondary user to diagnose primary user emulation attacks. Using a central-distributed hybrid approach ensures that our algorithm is highly adaptive, accurate and simple to implement.