Quantifying Geometry Effects on Low-Cost Intelligent Reflecting Surfaces

TL;DR

This paper evaluates how cost-saving measures like binary phase control and element grouping affect IRS performance in mmWave links, providing guidelines for practical deployment considering geometry and control complexity.

Contribution

It quantifies the performance trade-offs of simplified IRS architectures with binary control and grouping, offering deployment guidelines based on geometry.

Findings

Binary phase control reduces median SNR by ~4 dB.

Column-wise grouping introduces similar SNR penalties.

Even with constraints, large IRS can still significantly improve SNR.

Abstract

Intelligent Reflecting Surfaces (IRS) promise low-power coverage extension, yet practical deployments must curb hardware complexity and control overhead. This paper quantifies the performance impact of two cost-saving measures, column-wise element grouping and 1-bit (binary) phase quantization, relative to the ideal fully-controlled, continuous-phase baseline. A single-input single-output link is simulated at 26 GHz (mmWave) across three deployment geometries that vary the relative heights of access point, IRS and user equipment. Results show that switching from continuous to binary phase control reduces median SNR gain by approximately 4 dB, while adopting column-wise grouping introduces a similar penalty; combining both constraints incurs approximately 8 dB loss under height-offset deployments. When all nodes share the same height, the degradation from column-wise control becomes negligible, indicating deployment geometry can offset control-granularity limits. Despite the losses, a 32 x 32 column-wise binary IRS still delivers double-digit SNR gains over the no-IRS baseline in most positions, confirming its viability for cost-constrained scenarios. The study provides quantitative guidelines on when simplified IRS architectures can meet link-budget targets and where full element-wise control remains justified.