Resource Allocation Optimization in Power Wireless Private Networks based on Deep Reinforcement Learning

Abstract

Power wireless private networks are required to support reliable and low-latency communications for diverse power system applications under limited radio resources. Efficient resource allocation in such networks is challenging due to dynamic channel conditions, heterogeneous quality-of-service requirements, and the coupling among multiple resources. To address these challenges, this paper formulates the resource allocation problem as a sequential decision-making process and proposes a deep reinforcement learning-based optimization framework. By modeling network states, allocation actions, and performance-driven rewards, the proposed method enables adaptive and real-time resource allocation without relying on accurate system models. Simulation results demonstrate that the proposed approach significantly outperforms conventional allocation schemes in terms of system throughput and delay performance, while effectively satisfying quality-of-service constraints. These results indicate the effectiveness and robustness of deep reinforcement learning for resource management in power wireless private networks.