Impact of 5G Network Slicing on Efficiency and Stability Control of Base Station Microgrids

As power systems with multiple microgrids evolve to integrate distributed and renewable energy sources across different areas and regions, maintaining stable microgrid voltage and frequency control becomes increasingly reliant on fast and reliable communication infrastructure. Despite recent progress in wireless communication, many existing solutions still suffer from latency and packet loss that degrade control loop performance. This work presents a simulation framework that couple DC microgrids with proportional–integral (PI) voltage control loops and a 5G slicing-aware wireless communication model. A Q-learning-based algorithm is introduced to dynamically allocate bandwidth between two microgrids across three network slice types, namely legacy, shared, and ultra-reliable low-latency communication (URLLC) to optimize control performance under constrained radio resources. The impact of communication latency and reliability is evaluated through key performance indicators including voltage recovery, control reward, delay, packet loss, and throughput. Simulation results demonstrate that URLLC slicing significantly enhances control performance. Specifically, it reduces voltage recovery time by approximately 24%, lowers control-packet loss rate by up to 70% relative to legacy modes, and sustains a throughput improvement exceeding 20%. These results underscore the pivotal role of low-latency and high-reliability communication in enabling stable and efficient smart grid operation.