Subset Antenna Selection for Simultaneously Transmitting and Reflecting Surface Assisted Wireless Systems

Abstract

Simultaneously transmitting and reflecting recon-figurable intelligent surfaces (STAR-RIS) enable full space coverage in a cost and energy-efficient manner. Subset antenna selection (AS), where the transmitter selects a subset of antennas based on channel conditions, harnesses diversity benefits at the cost of fewer radio frequency (RF) chains. To combine these benefits, we consider subset AS for a STAR-RIS-assisted system under independent and coupled phase shift models. We develop an algorithm that jointly optimizes the antenna subset, access point beamforming vectors, and the STAR-RIS transmission and reflection coefficients to maximize the rate achieved by the T-user subject to a minimum rate constraint for the R-user. It performs an exhaustive search over all possible subsets and uses alternating optimization to obtain beamforming vectors and STAR-RIS phase shifts. We also propose a low-complexity subset selection algorithm based on a metric derived from direct and cascaded link channel gains. We evaluate the impact of the STAR-RIS phase shift model on the achievable T-user rate and the $\mathbf{R}$-user outage probability. We also show that the T-user rate increases with the number of STAR-RIS elements or RF chains.