Budgeted Thompson Sampling for IRS Enabled WiGig Relaying

Intelligent reconfigurable surface (IRS) is a competitive relaying technology to widen the WiGig coverage range, as it offers an effective means of addressing blocking issues. However, selecting optimal IRS relay for maximum attainable data rate time-consuming process, requires beamforming training (BT) tune phase shifts (PSs) base station (WGBS) and relays. This paper proposes self-learning-based budgeted Thomson sampling approach probing (BTS-IRS) address this challenge. The BT time cost incorporated into main BTS formula, where both payoff posterior distributions are sampled separately, their ratio estimated, arm/IRS with highest decided. enables be chosen lowest cost. Numerical results demonstrate improved performance BTS-IRS technique regarding consumption/cost, spectral efficiency, when compared other benchmarks.