Early Acceptance Matching Game for User-Centric Clustering in Scalable Cell-free MIMO Networks

TL;DR

This paper introduces an early acceptance matching algorithm for user-centric clustering in scalable cell-free MIMO networks, reducing signaling and improving QoS satisfaction compared to existing methods.

Contribution

It proposes a novel early acceptance matching game for dynamic AP-UE clustering, enhancing scalability and efficiency in cell-free MIMO networks.

Findings

Reduces fronthaul signaling significantly.

Increases the number of UEs meeting QoS requirements.

Outperforms state-of-the-art clustering approaches.

Abstract

The canonical setup is the primary approach adopted in cell-free multiple-input multiple-output (MIMO) networks, in which all access points (APs) jointly serve every user equipment (UE). This approach is not scalable in terms of computational complexity and fronthaul signaling becoming impractical in large networks. This work adopts a user-centric approach, a scalable alternative in which only a set of preferred APs jointly serve a UE. Forming the optimal cluster of APs for each UE is a challenging task, especially, when it needs to be dynamically adjusted to meet the quality of service (QoS) requirements of the UE. This complexity is even exacerbated when considering the constrained fronthaul capacity of the UE and the AP. We solve this problem with a novel many-to-many matching game. More specifically, we devise an early acceptance matching algorithm, which immediately admits or rejects UEs based on their requests and available radio resources. The proposed solution significantly reduces the fronthaul signaling while satisfying the maximum of UEs in terms of requested QoS compared to state-of-the-art approaches.