A flexible multi-criteria scheme to detect primary user emulation attacks in CRAHNs

Cognitive Radio Ad Hoc Networks (CRAHNs) are prone to Primary User Emulation Attacks (PUEAs), a particular security issue in which malicious unlicensed users manipulate their cognitive radio parameter values to monopolize the spectrum usage pretending transmissions of Primary Users. Works in the literature propose to detect or mitigate these attacks, however they are based on network architectural approaches and node operation, without intending to be flexible to unexpected attack behaviors and prepared for attack adaptations that intend to circumvent defense mechanisms. Existing proposals are intrinsically tied to predefined criteria, and the addition of new one requires the design of a different solution. Intending to fill this gap, we propose INCA, a novel multIple criteria scheme for a deceNtralized and Cooperative Analysis of the PUEA presence in CRAHNs. INCA is composed of two phases that employ, respectively, the Normalized Weighted Additive Utility Function to achieve flexibility, and the Bayes theorem to provide a distributed cooperation among nodes, in which the detection of the presence of attacks by a node is partially conditional to the detection by other nodes. Results from exhaustive simulations show that INCA can improve the analysis of the presence of PUEA in all evaluated scenarios, mostly when both phases are applied.