Data-Driven Learning of Two-Stage Beamformers in Passive IRS-Assisted Systems with Inexact Oracles

TL;DR

This paper presents a novel data-driven, model-free unsupervised learning algorithm for passive IRS-assisted beamforming that operates without channel knowledge, using zeroth-order stochastic gradient ascent to handle inexact evaluations in complex wireless systems.

Contribution

It introduces a new zeroth-order stochastic gradient ascent method for two-stage IRS beamforming, capable of handling inexact oracles and operating without channel models or statistics.

Findings

Effective in large-scale MISO downlink scenarios

Operates without channel model assumptions

Converges close to stationary points

Abstract

We develop an efficient data-driven and model-free unsupervised learning algorithm for achieving fully passive intelligent reflective surface (IRS)-assisted optimal short/long-term beamforming in wireless communication networks. The proposed algorithm is based on a zeroth-order stochastic gradient ascent methodology, suitable for tackling two-stage stochastic nonconvex optimization problems with continuous uncertainty and unknown (or "black-box") terms present in the objective function, via the utilization of inexact evaluation oracles. We showcase that the algorithm can operate under realistic and general assumptions, and establish its convergence rate close to some stationary point of the associated two-stage (i.e., short/long-term) problem, particularly in cases where the second-stage (i.e., short-term) beamforming problem (e.g., transmit precoding) is solved inexactly using an arbitrary (inexact) oracle. The proposed algorithm is applicable on a wide variety of IRS-assisted optimal beamforming settings, while also being able to operate without (cascaded) channel model assumptions or knowledge of channel statistics, and over arbitrary IRS physical configurations; thus, no active sensing capability at the IRS(s) is needed. Our algorithm is numerically demonstrated to be very effective in a range of experiments pertaining to a well-studied MISO downlink model, including scenarios demanding physical IRS tuning (e.g., directly through varactor capacitances), even in large-scale regimes.