Intelligent Reflecting Surface-based Secrecy Rate Enhancement in Multicast Multigroup Tactical Communication Systems

Abstract

Improving wireless transmission capacity and security is substantial in today's tactical networks due to the explosion of the data-intensive applications. In this paper, we investigate a framework that can successfully convey the messages from the transmitter to the receivers with the supports of intelligent reflecting surfaces (IRSs) to achieve high data rate. In particular, we manage the actual transmitted data instead of sending all source data in IRSs-aided multicast multigroup systems. A key challenging issue of this problem consists in obtaining a closed-form expression for the complicated distribution of the signal-to-noise-plus-interference (SINR) in IRSs-aided wireless systems. Therefore, we propose a deep neural network (DNN)-based framework to obtain a high-accuracy prediction of the long-term network capacity. Based on predicted results, we solve the joint power control and data reduction ratio selection problem to maximize the total received quality of service (QoS). Furthermore, we also consider the physical layer security design where data leakage among users' groups is prevented. We adapt well-known zero-forcing (BD) and block diagonalization (BD) techniques in achieving high-efficient secure solutions in IRSs-aided multicast multi-group systems. Numerical results confirm the efficiency of the proposed design. The information leakage can decrease down to 0 thanks to our proposed framework.