Back to search

University of Sheffield

Resource allocation for user-centric cell-free massive MIMO under hotspot traffic and mixed LoS/NLoS propagation

Abstract

dc:description.abstract

This research investigates pilot assignment and Access Point (AP) selection for user-centric Cell-Free Massive Multiple-Input Multiple-Output (CF-mMIMO) networks, where pilot reuse can contaminate channel estimates and reduce uplink spectral efficiency (SE). Two settings are considered separately: hotspot traffic under Rayleigh fading and closely spaced users under mixed Line-of-Sight (LoS)/Non-Line-of-Sight (NLoS) propagation. The objective is to reduce pilot contamination and improve average uplink SE and 95%-likely SE (the SE exceeded by 95% of users) while controlling allocation cost and respecting service-capacity constraints. Hotspot geometry guides hierarchical pilot assignment, while array-relative directions and realized LoS states guide joint AP selection and pilot reuse. This research proposes a hotspot-aware pilot-assignment method that combines canopy-assisted k-means (C-KM), local graph-partitioning-based grouping, and Hungarian mapping. Clustering uses user positions and estimated serving extent to limit pairwise processing to local groups, where graph partitioning separates strongly coupled user pairs into different pilot groups when possible. Hungarian mapping uses normalized serving-AP signatures to coordinate pilot reuse across clusters. Simulation results show lower normalized mean square error and a modest gain in 95%-likely SE over global graph partitioning. Total pilot-assignment time, including C-KM clustering, is also lower, with savings depending on cluster balance. For mixed LoS/NLoS propagation, this research proposes Geometry and LoS Aware Variable Neighbourhood Descent (GLoS-VND) for joint AP selection and pilot assignment. Given user positions, array orientations, LoS states, and channel statistics, the method evaluates candidate allocations using a common objective based on retained estimated channel energy, accounting for signal strength and spatial overlap between users sharing a pilot. Neighbourhood descent searches feasible changes to AP associations and pilot assignments, accepting moves that improve this objective. Simulation results show higher average uplink SE than the evaluated reference methods, with improvements in 95%-likely SE at higher user loads.

Degree

thesis:*
Name dc:type.qualificationname
Ph.D
Level dc:type.qualificationlevel
doctoral
Grantor dc:publisher.institution
University of Sheffield
Year dc:date.issued
2026

Author and committee

dc:creator, dc:contributor.*
Author dc:creator
  • Yang, Hansong
Advisor dc:contributor.advisor
  • Chu, Xiaoli

Rights

Language dc:language.dcterms_iso639_2
eng

Chain of custody

source
Harvested from
White Rose University Consortium
Base URL
etheses.whiterose.ac.uk/cgi/oai2
Last updated
2026-11-19
Source record
OAI-PMH GetRecord
related terms
citation

Yang, Hansong. Resource allocation for user-centric cell-free massive MIMO under hotspot traffic and mixed LoS/NLoS propagation. doctoral thesis, University of Sheffield, 2026.