Resource Load Balancing Algorithm for Mobile Edge Computing System Based on Deep Reinforcement Learning

Due to the complex and variable combat environment, high requirements for data transmission confidentiality, and the imperfections in centralized data center construction, the application of cloud computing is significantly limited, failing to meet the massive data processing demands. Utilizing mobile edge computing, which processes tasks near its source, can effectively address this issue, demonstrating promising application prospects. Load balancing is a key technology for ensuring high availability and quality of service (QoS) in Mobile Edge Compu-ting (MEC) services. This paper designs a resource load balancing algorithm for edge servers based on the Proximal Policy Optimization (PPO) algorithm, considering CPU resources, storage resources, and network resources as the load measurement indicators of the edge system. The algorithm schedules tasks within the system to achieve the goal of minimizing the average load of the mobile edge computing system. Simulation experiments have verified that the algorithm pro-posed in this paper can achieve a lower average system load more quickly compared to three other baseline algorithms.