Site-Specific Deployment Optimization of Intelligent Reflecting Surface for Coverage Enhancement

TL;DR

This paper investigates site-specific deployment of IRS in wireless networks, considering antenna patterns, and demonstrates that active IRSs significantly improve coverage and fairness compared to passive IRSs through optimized placement and ILP solutions.

Contribution

It introduces a site-specific deployment framework for IRS considering antenna patterns and optimizes placement using ILP, highlighting the advantages of active IRSs over passive ones.

Findings

Active IRSs outperform passive IRSs in coverage and fairness.

Antenna and IRS element patterns critically affect link performance.

ILP-based deployment optimization is effective for moderate-sized networks.

Abstract

Intelligent Reflecting Surface (IRS) is a promising technology for next generation wireless networks. Despite substantial research in IRS-aided communications, the assumed antenna and channel models are typically simplified without considering site-specific characteristics, which in turn critically affect the IRS deployment and performance in a given environment. In this paper, we first investigate the link-level performance of active or passive IRS taking into account the IRS element radiation pattern (ERP) as well as the antenna radiation pattern of the access point (AP). Then the network-level coverage performance is evaluated/optimized in site-specific multi-building scenarios, by properly deploying multiple IRSs on candidate building facets to serve a given set of users or Points of Interests (PoIs). The problem is reduced to an integer linear programming (ILP) based on given link-level metrics, which is then solved efficiently under moderate network sizes. Numerical results confirm the impact of AP antenna/IRS element pattern on the link-level performance. In addition, it is found that active IRSs, though associated with higher hardware complexity and cost, significantly improve the site-specific network coverage performance in terms of average ergodic rate and fairness among the PoIs as well as the range of serving area, compared with passive IRSs that have a much larger number of elements.