Secrecy Performance Enhancement Using Path Selection over Cluster-Based Cognitive Radio Networks

In this paper, we propose three path selection methods for cluster-based cognitive radio (CR) networks for secrecy enhancement by formulating the probability of non-zero secrecy capacity (PNSC). In the proposed work, it is assumed that uniform transmit power for the secondary transmitters and jammers must be adjusted to guarantee quality of service (QoS) of the primary network, follows a simple and efficient power allocation strategy. To improve the channel capacity, the best receiver is selected at each cluster to relay the source data to the next hop. Additionally, a jammer is randomly chosen at each cluster to generate noises on an eavesdropper, and to reduce the quality of the eavesdropping links. Three methods are studied in this paper. First, we propose the BEST path selection method (BEST) to maximize the end-to-end instantaneous secrecy capacity. Second, the path obtaining the MAXimum Value for the average end-to-end PNSC (MAXV) is selected for data transmission. Third, we also propose a RAND method in which a RANDom path is employed. For performance evaluation and comparison, we derive exact closed-form expressions for the end-to-end PNSC of the BEST, MAXV and RAND methods over Rayleigh fading channel. Monte Carlo simulations are then performed to verify the derived theoretical results.