Improved detection of primary user emulation attacks in cognitive radio networks

Abstract

Modern society heavily depends on wireless spectrum for communication purposes. With the rapid increase in mobile devices and Internet of Things (IoT), the need for wireless spectrum has grown dramatically resulting in the limited available spectrum becoming a constrained resource. To remedy this spectrum scarcity, cognitive radio (CR) was proposed as an efficient and opportunistic use of the frequency spectrum in order to increase spectral efficiency. However dynamic sharing of the spectrum between multiple users poses several significant challenges in security and trust. In this paper, we focus on primary user emulation (PUE) attacks in cognitive radio networks (CRNs). We propose a new scheme that we believe could be useful in practice to achieve an improved PUE attack detection in CRNs. The scheme combines energy detection and localization. One of the distinguishing features of the proposed scheme is that instead of detecting received energy level with a single threshold for a secondary user (SU), multiple thresholds have been used for each SU and the global decision is concluded by the majority of participating SUs. Furthermore, in cases where the primary users (PUs) are stationary and their coordinates are already known to SUs, we have combined our improved energy detection with a TDOA localization scheme for detecting PUE attacks. Our simulation results show that this scheme increases the level of accuracy in detecting PUE attacks.