Placement Optimization and Power Control in Intelligent Reflecting Surface Aided Multiuser System

TL;DR

This paper proposes a joint IRS placement and power control scheme to maximize the minimum throughput in multiuser systems, demonstrating significant performance improvements over benchmarks through a novel ring-based placement and efficient algorithm.

Contribution

It introduces a joint multi-IRS placement and power control framework with a ring-based scheme and an efficient search algorithm for multi-IRS multiuser systems.

Findings

Proposed scheme outperforms benchmarks in throughput.

IRSs are best placed near AP for coverage, spreading out with more IRSs.

Efficient algorithm achieves near-optimal solutions for various IRS configurations.

Abstract

Intelligent reflecting surface (IRS) is a new and revolutionary technology capable of reconfiguring the wireless propagation environment by controlling its massive low-cost passive reflecting elements. Different from prior works that focus on optimizing IRS reflection coefficients or single-IRS placement, we aim to maximize the minimum throughput of a single-cell multiuser system aided by multiple IRSs, by joint multi-IRS placement and power control at the access point (AP), which is a mixed-integer non-convex problem with drastically increased complexity with the number of IRSs/users. To tackle this challenge, a ring-based IRS placement scheme is proposed along with a power control policy that equalizes the users' non-outage probability. An efficient searching algorithm is further proposed to obtain a close-to-optimal solution for arbitrary number of IRSs/rings. Numerical results validate our analysis and show that our proposed scheme significantly outperforms the benchmark schemes without IRS and/or with other power control policies. Moreover, it is shown that the IRSs are preferably deployed near AP for coverage range extension, while with more IRSs, they tend to spread out over the cell to cover more and get closer to target users.