Power Efficient Robust Beamforming Design for Active RIS-Aided MU-MIMO SWIPT Systems

This paper investigates an unconventional framework of active reconfigurable intelligent surface (ARIS) aided multi-user (MU) multi-input multi-output (MIMO) system for simultaneous wireless information and power transfer (SWIPT) network under statistical channel estimation error (CEE). Particularly, we focus on the problem of power minimization via joint beamforming design at BS and ARIS for considered SWIPT system while guaranteeing the minimum rate requirement and the minimum energy-harvested constraints for information and energy receivers, respectively. Owing to the non-convex and NP-hard nature of the formulated problem, we first utilize a minimum mean square error (MMSE) approach to transform the problem into its simplified form, and later utilize an alternating optimization framework which solves the problems of beamforming design at BS and ARIS independently in an iterative manner using general approximations. Simulation results confirm that ARIS can significantly reduce the required transmission power by 50-60% when compared to passive RIS while satisfying given QoS constraints for SWIPT systems under the CEE model.

I. Introduction

Simultaneous wireless information and power transfer (SWIPT) systems have been considered a new paradigm for green communication architecture in next-generation wireless networks [1], [2]. The smart energy harvesting inherited in multi-antenna SWIPT system renders the better quality of services to a diverse set of heterogeneous users in terms of energy harvesting and information transmission capacity, which in turn become a paramount interest for energy-constrained future networks [3]. However, the performance of multi-user (MU) multi-input-multi-output (MIMO) SWIPT systems is sensitive to poor or uncertain channel conditions [4].