Impact of Sociality Regimes on Quality of Service and Energy Efficiency in Cell-Free MIMO Networks

TL;DR

This paper explores how different social cooperation regimes among users and access points in cell-free MIMO networks affect the balance between quality of service and energy efficiency, proposing novel matching algorithms.

Contribution

It introduces a social matching game model with externalities for optimizing AP-UE clustering considering sociality regimes and proposes two new algorithms based on Deferred Acceptance and Early Acceptance.

Findings

Egalitarian sociality regime yields the best QoS and energy efficiency trade-off.

Sociality regimes significantly influence network performance metrics.

Proposed algorithms effectively balance QoS and energy consumption.

Abstract

The cell-free architecture represents a significant advancement in network design, where each User Equipment (UE) is served by a group of distributed Access Points (APs), aimed at delivering uniformly high data rates to UEs across all locations. To ensure network scalability, user-centric clustering (UCC) has emerged as a practical approach, wherein only a selected subset of preferred APs jointly serves each UE. Forming the optimal cluster of APs for each UE is a challenging task, particularly when limited fronthaul and processing capabilities of both APs and UEs are considered. This challenge is exacerbated by the need for dynamic adjustments to enhance energy efficiency while meeting quality of service (QoS) requirements, which introduces conflicting interests between these entities. In this paper, we investigate the sociality regime of UEs and their clusters of APs, characterizing their intra-team cooperation as either selfish, egalitarian, or altruistic. These sociality regimes are crucial in achieving a balance between QoS and energy efficiency per UE. We address this problem by modeling it as a many-to-many social matching game with externalities, where connections can be established based on the sociality regime of both teams. To solve this, we introduce two novel algorithms based on Deferred Acceptance (DA) and Early Acceptance (EA) matching games. Numerical results show the significant impact of the sociality regime adopted by UEs and their clusters of APs on UE's QoS satisfaction and energy efficiency, with the egalitarian regime adopted by both entities proving the best performance trade-off.