Joint malicious user detection and resource allocation in cognitive radio networks

Cognitive radio is an effective technology to solve the increasing spectrum demands for wireless application by reducing the spectrum based on the sensing results. However, the reliability of each secondary user (SU) is uncertain. In order to improve the chance to access the idle spectrum, some SUs send false sensing results and become malicious users (MUs). Our objective is to design a novel algorithm that joint MUs detection and power allocation to optimize the total utility of the whole cognitive radio network, while meets the requirement of MUs detection accuracy as well as the interference limitation to primary users (PUs). The utility is composed of three parts: the detection probability of MUs, sum-rate and the energy consumption for MUs detection. We formulate the problem into two sub-problems. Firstly, we detect MUs to get the detection probability of MUs and the energy consumption for MUs detection. Sencondly, based on the results from the first step we use the Lagrange dual decomposition method to maximize the sum-rate in the network.