Multi-Cell MMSE Processing and Pilot Assignment in Massive MIMO under Correlated Fading

Elyasi, Masoud; Vosoughi, Azadeh

Permalink

https://hdl.handle.net/2142/130295 Copy

Description

Title

Multi-Cell MMSE Processing and Pilot Assignment in Massive MIMO under Correlated Fading

Author(s)

• Elyasi, Masoud
• Vosoughi, Azadeh

Issue Date

2025-09-17

Keyword(s)

• Massive MIMO
• Spatially correlated Rayleigh fading
• Multi-cell processing
• M-MMSE
• P-MMSE
• Pilot contamination
• Scalable pilot assignment
• Large system analysis
• Random matrix theory

Abstract

This paper investigates a multi-cell massive MIMO network subject to spatially correlated Rayleigh fading, where pilot reuse occurs both within and across cells. Base stations (BSs) implement multi-cell minimum mean square error (M-MMSE) processing to mitigate interference. We present a new deterministic approximation of the uplink coherent interference term, which remains accurate in the large-antenna regime despite the presence of spatial correlation and pilot reuse—addressing a critical gap in prior analyses. Based on this analytical result, we design a multi-cell pilot assignment (PA) algorithm that fully removes pilot contamination by utilizing the spatial correlation information of all users. To enable deployment in large-scale networks, we also develop a scalable variant using partial M-MMSE (P-MMSE) processing, which restricts inter-BS coordination by selectively engaging only strong interferers, and provide a complexity evaluation to demonstrate its implementation feasibility. Simulation results confirm that our proposed PA algorithms significantly outperform existing PA schemes across various system configurations, achieving higher spectral efficiency, improved fairness, and reduced pilot overhead—offering practical design insights for next-generation massive MIMO systems.

Publisher

Allerton Conference on Communication, Control, and Computing

Series/Report Name or Number

2025 61st Allerton Conference on Communication, Control, and Computing Proceedings

ISSN

2836-4503

Type of Resource

Text

Genre of Resource

Conference Paper/Presentation

Language

eng

Handle URL

https://hdl.handle.net/2142/130295&&

Copyright and License Information

Copyright 2025 is held by Masoud Elyasi and Azadeh Vosoughi.

Owning Collections