Enabling Seamless Connectivity in Consumer Electronics: A DRL Approach for Scalable Cell-Free mMIMO Handover under O-RAN

Scalability challenges in modern networks arise from the increasing density of User Equipment (UE) and Access Points (APs), compounded by hardware heterogeneity and complexity. Cell-Free Massive MIMO (CF mMIMO) addresses these issues by adopting a UE-centric paradigm, supported by the flexible O-RAN architecture. However, the integration of these technologies in the context of UE mobility requires further investigation. This study optimizes CF mMIMO deployment in O-RAN by proposing a Deep Reinforcement Learning (DRL) approach to maximize throughput while balancing the competing constraints of latency and handover costs. Compared to cellular, fixed, and ubiquitous baselines, the DRL model achieves 30% and 45% higher spectral efficiency, respectively. Furthermore, a 22% reduction in unnecessary handovers has been achieved by deploying the proposed DRL model. The results highlight significant implications for consumer electronics, enabling reliable connectivity in dense Internet of Things (IoT) and Augmented Reality (AR) environments through enhanced Quality of Experience (QoE).