Deployment and adaptive optimization strategies for B5G heterogeneous edge base station microgrid using K-means and MPC

The exponential growth in smart devices and mobile data traffic presents significant challenges for Beyond 5G (B5G) networks. The dense deployment of these networks has escalated energy consumption, conflicting with existing goals for energy conservation and emission reduction. Base Station Microgrids (BSMGs), powered by renewable energy, offer a promising solution by alleviating energy pressure on operators due to their economic and environmental advantages. However, recent research on base station deployment has mainly concentrated on performance and coverage, often neglecting the costs to communication operators. Furthermore, the massive increase in device access within heterogeneous networks necessitates urgent improvements in network capacity and the optimization of computing resources using edge computing. This paper proposes a B5G heterogeneous edge BSMG system, comprising macro BSMGs and edge BSMGs. An enhanced K-means algorithm is employed to optimize the deployment of BSMGs, while an adaptive optimization strategy, incorporating edge computing and model predictive control, is designed to maximize green energy utilization. Extensive simulations demonstrate that the proposed system effectively reduces reliance on the traditional power grid and optimizes energy and computing resources compared to other schemes.