Efficient design of reconfigurable intelligent surface assisted multi-group multicast beamforming

Abstract

This thesis considers a multi-group multicasting scenario facilitated by a reconfigurable intelligent surface (RIS). We propose low-complexity scalable algorithms for the joint design of the base station (BS) multicast beamformer and the RIS passive beamformer to minimize the transmit power subject to the quality-of-service (QoS) constraints and maximize the minimum quality-of-service among all users mainting a power budget constraint which is also known as max-min fair (MMF) problem. By exploring the interaction of the BS and RIS beamforming in the QoS problem, we formulate two subproblems, a BS multicast QoS problem and a RIS max-min-fair multicast problem, to be solved alternatingly. Our alternating multicast beamforming approach (AMBA) not only enables us to exploit the optimal multicast beamforming structure at the BS, but also allows us to employ a fast first-order projected subgradient algorithm (PSA) to solve the RIS MMF subproblem. Furthermore, we show that QoS problem and MMF problem in our RIS-aided scenario are inverse problems, and by using the optimal beamforming structure for BS beamformer, we propose a low complexity and efficient algorithm to solve MMF problem as well. The mentioned algorithm first reformulate the MMF problem by turning the two objective parameters in MMF problem (BS and RIS beamforming vectors) into a single one and then employs fast first-order projected subgradient algorithm here, and at the same time avoids usual alternating process which is commonly used for two-objective parameters problems. After analysing the multicast scenario, we will show that our algorithms is applicable for general case of unicast as well. Simulation results show the effectiveness of our proposed alternating approach for QoS problem and its advantage in terms of both performance and computational complexity over other alternative methods. Further simulations show the behavior of our PSA-based algorithm for MMF problem for different set-up parameters, and its effectiveness over similar proposed methods in special cases.