Monitoring of Carbon Footprint and Traffic Optimization in 5G Network

Abstract

The expansion of 5G infrastructure and the impending evolution to 6G technology bring forward significant sustainability challenges due to increased energy demands, particularly within the Radio Access Network (RAN). This study examines the potential of Artificial Intelligence (AI) to support energy-efficient operations in next- generation wireless networks. We introduce AI4GreenNet, a novel AI-based framework that leverages deep reinforcement learning and graph neural networks to streamline network functionality, minimize energy-intensive tasks, and manage power usage at base stations. The framework is designed to work seamlessly within Open RAN (O-RAN) environments, enabling intelligent, vendor-agnostic optimization of disaggregated RAN components. By dynamically adapting to traffic loads and energy availability, the model demonstrates substantial improvements in energy savings and carbon emission reductions. Simulations indicate up to 45% reduction in carbon emissions in urban deployments, aligning future wireless connectivity with global sustainability goals.