Latency Optimization in Clustering NOMA-Aided Cell-Free Massive MIMO with Mobile Edge Computing

Reducing network latency while maintaining energy efficiency is a key challenge in next-generation wireless networks, particularly in a clustering NOMA-aided cell-free massive multiple-input multiple-output (CF-mMIMO) system with mobile edge computing (MEC). Unlike conventional pairing-based NOMA, we propose a novel clustering NOMA approach using Kmeans algorithm, which efficiently groups user equipments (UEs) into clusters based on their channel conditions and proximity to access points (APs). The objective is to optimize power allocation and offloading ratios while minimizing the overall system latency and maintaining energy efficiency. To tackle this nonconvex optimization problem, we introduce a two-step solution framework, where the first step applies K-means clustering for optimal UE grouping, and the second step employs successive convex approximation (SCA) to jointly optimize power allocation and offloading decisions. Simulation results demonstrate that our proposed clustering-based NOMA approach for MEC-enabled CF-mMIMO networks achieves faster convergence, lower latency, and higher energy efficiency compared to traditional pairingbased NOMA, CF-mMIMO, and conventional massive MIMO schemes.