Cellular Network Radio Monitoring and Management through Virtual UE Probes: A Study Case Based on Crowded Events

Abstract

Although log processing of network equipment is a common technique in cellular network management, several factors make said task challenging, especially during mass attendance events. The present paper assesses classic methods for cellular network measurement and acquisition, showing how the use of on-the-field user probes can provide relevant capabilities to the analysis of cellular network performance. Therefore, a framework for the deployment of this kind of system is proposed here based on the development of a new hardware virtualization platform with radio frequency capabilities. Accordingly, an analysis of the characteristics and requirements for the use of virtual probes was performed. Moreover, the impact that social events (e.g., sports matches) may have on the service provision was evaluated through a real cellular scenario. For this purpose, a long-term measurement study during crowded events (i.e., football matches) in a stadium has been conducted, and the performances of different services and operators under realistic settings has been evaluated. As a result, several considerations are presented that can be used for better management of future networks.

Beyond REM: A New Approach to the Use of Image Classifiers for the Management of 6G Networks

“…In practical deployment, a feedback loop can be established for continuous model updating with fresh data, enabling real-time optimization and adaptation to changing network conditions. This process, coupled with an integrated approach to data collection [31], diagnosis, and action, can effectively facilitate network self-healing. Considering the diversity and complexity of network issues, continuous training with new problem instances is necessary to comprehensively cover all potential faults and their variations, which might require regular model updates using newly collected data, the frequency of which depends on the network size, its rate of change, and available computational resources.…”