Joint Antenna Selection and Beamforming for an IRS Aided IoT System

Adit Jain, Rimalapudi Sarvendranath, Salil Kashyap

Abstract: Intelligent reflecting surface (IRS), which is made up of passive reflective elements and can control the phase of the incident signal, and antenna selection (AS) can be combined to yield a cost-and energy-efficient wireless technology for the Internet of Things (IoT) system. For an IRS-assisted IoT system with one fusion node and multiple sensor nodes, we develop a jointly optimal AS and passive beamforming rule that maximizes the sum data rate. In it, the number of required channel estimations increases linearly with the number of sensor nodes. Additional novel contributions include a closed-form AS and passive beamforming rule, which maximizes the sum of absolutes of channel gains while significantly reducing computational complexity. To further simplify, we propose a new channel acquisition procedure for which the number of channel estimations is independent of the number of sensor nodes. Our simulations show that the optimal rule yields up to 13.6 x and 6 x higher rates than the maximum channel gain based and block coordinate descent based algorithms, respectively. Furthermore, they show that the simpler AS rule yields up to 12.4x gain compared to other AS rules in the literature and is robust to estimation errors.