A comparative performance analysis of phase configuration algorithms in Near-Field for Multi-User LIS-Aided UAV systems

Abstract

Large intelligent surfaces (LIS) are a key 6G
technology, but extremely large antenna arrays (ELAA) necessitate
near-field (NF) operation, challenging conventional farfield
(FF) beamforming. This paper addresses NF multi-user
(MU) beamforming in a LIS-aided Unmanned Aerial Vehicle
(UAV) relay system. We develop and compare two distinct phase
configuration algorithms for multi-beam NF focusing: a lowcomplexity
Grouped Beamforming heuristic and a Grey Wolf
Optimizer (GWO) Beamforming metaheuristic. Performance,
including Average Array Gain, Average Minimum Array Gain,
their Cumulative Distribution Functions, and computational
complexity, is evaluated via Monte Carlo simulations against
baseline methods. Both proposed algorithms significantly outperform
baselines. GWO Beamforming especially ensuring fairness
under high user loads, but incurs a considerable computational
cost. Conversely, Grouped Beamforming offers a practical realtime
solution. Our findings illuminate the critical performancecomplexity
trade-offs guiding algorithm selection for NF MU
LIS-UAV deployments.