Machine Learning-Based Coverage and Capacity Optimization xApp/rApp for Open RAN 5G Campus Networks

Abstract:
The Open Radio Access Network (O-RAN) paradigm is gaining ever-increasing attention and is expected to revolutionize the telecommunications market. It promotes virtualized and disaggregated network components interconnected via open interfaces within multi-vendor environments. Optimization in O-RAN is achieved via RAN intelligent Controllers (RICs) employing data-driven and automated processes known as rApps and xApps. However, achieving optimized network performance in O-RAN poses several challenges, prompting consideration of various use cases. One of the major use cases and the main focus of this paper is Coverage and Capacity Optimization (CCO) which aims to ensure comprehensive coverage in ORAN deployments without excessive costs. This paper presents a novel self-organized machine learning-based CCO-xApp/rApp, leveraging Recommender Systems (RecSys), to address coverage and capacity issues from a global network management perspective. The introduced CCO-xApp/rApp has been deployed on an open-source O-RAN-compliant testbed. Our results show that the CCO-xApp/rApp can autonomously optimize network performance. To the best of our knowledge, this represents the first successful demonstration of such an xApp/rApp, marking a pioneering achievement in this area.