User-Centric Cell-Free Massive MIMO Enhanced by Fluid-Antenna Access Points: Uplink Analysis

TL;DR

This paper introduces a user-centric cell-free massive MIMO system with fluid antennas at access points, proposing novel channel estimation, port selection, and optimization techniques that enhance spectral efficiency and system scalability.

Contribution

It develops a comprehensive framework for fluid-antenna-based CF-mMIMO, including dynamic port activation, spatial correlation exploitation, and an optimization method for uplink spectral efficiency.

Findings

Fluid antennas significantly reduce channel estimation error.

Optimized port selection improves uplink sum spectral efficiency.

Proposed methods outperform fixed-antenna and non-optimized baselines.

Abstract

In this paper, we investigate cell-free massive MIMO (CF-mMIMO) systems in which access points (APs) are equipped with fluid antennas (FAs) and develop a comprehensive framework for channel estimation, antenna port selection, and uplink spectral efficiency (SE) optimization. We propose a generalized LMMSE-based uplink channel estimation scheme that dynamically activates FA ports during pilot transmission, efficiently exploiting antenna reconfigurability under practical training constraints. Building on this, we design a distributed port selection strategy that minimizes per-AP channel estimation error by exploiting spatial correlation among FA ports. We systematically analyze the impact of antenna geometry and spatial correlation using the Jakes' channel model for different AP array configurations, including uniform linear and planar arrays. We then derive SINR expressions for centralized and distributed uplink processing and obtain a closed-form uplink SE expression for centralized maximum-ratio combining using the use-and-then-forget bound. Finally, we propose an alternating-optimization framework to select FA port configurations that maximize the uplink sum SE. Numerical results show that the proposed FA-aware channel estimation and port optimization strategies greatly reduce channel estimation error and significantly improve sum-SE over fixed-antenna and non-optimized FA baselines, confirming FAs as a key enabler for scalable, adaptive CF-mMIMO networks.