Near-Field Modelling and Performance Analysis of Modular Extremely Large-Scale Array Communications

TL;DR

This paper introduces a modular XL-array architecture for large-scale antenna systems, providing a detailed near-field model and analyzing its performance, including SNR scaling laws and the importance of near-field effects.

Contribution

It develops a novel near-field modelling framework for modular XL-arrays and derives a closed-form SNR expression, extending existing far-field models and revealing new scaling laws.

Findings

Near-field effects significantly impact array performance.

Derived a closed-form maximum SNR expression.

Established asymptotic SNR scaling law for large arrays.

Abstract

This letter studies a new array architecture, termed as modular extremely large-scale array (XL-array), for which a large number of array elements are arranged in a modular manner. Each module consists of a moderate number of array elements and the modules are regularly arranged with the inter-module space typically much larger than signal wavelength to cater to the actual mounting structure. We study the mathematical modelling and conduct the performance analysis for modular XL-array communications, by considering the non-uniform spherical wave (NUSW) characteristic that is more suitable than the conventional uniform plane wave (UPW) assumption for physically large arrays. A closed-form expression is derived for the maximum signal-to-noise ratio (SNR) in terms of the geometries of the modular XL-array, including the total array size and module separation, as well as the user's location. The asymptotic SNR scaling law is revealed as the size of modular array goes to infinity. Furthermore, we show that the developed modelling and performance analysis include the existing results for collocated XL-array or far-field UPW assumption as special cases. Numerical results demonstrate the importance of near-field modelling for modular XL-array communications since it leads to significantly different results from the conventional far-field UPW modelling.