Sum Secrecy Rate Maximization for Relay-Aided Multiple-Source Multiple-Destination Networks

This paper studies a multiple-source multiple-destination network with the presence of multiple eavesdroppers, in which an amplify-and-forward (AF) relay is used to bridge the communication between the source-destination pairs to overcome the long-distance attenuation. Considering the physical-layer security issues, we aim to maximize the sum secrecy rate subject to the relay power constraint and the quality-of-service (QoS) requirements for legitimate user equipment. First, we propose an algorithm based on the monotonic optimization and the semidefinite programming (MO-SDP). Simulation results show that our proposed MO-SDP algorithm exhibits almost the same performance as the optimal solution. To alleviate the problem of high complexity associated with the MO-SDP algorithm, we then propose an alternative solution based on the null-space (NuS) relay precoding, the complexity of which is significantly reduced, and it yields a semiclosed-form expression for the solution. Moreover, the performance of the proposed NuS algorithm is evaluated via simulations, and the performance of the NuS algorithm and that of the MO-SDP algorithm are shown to converge at the high signal-to-noise ratio (SNR) region.