A Cooperative Edge Offloading Strategy for New Power System Services

Multi-access edge computing (MEC) is a critical technology for 5G networks. Computing tasks can be processed in edge servers instead of cloud servers by deploying computing servers at the edge of integrated power communication networks. With the expanding scale of the power grid, the delay-sensitive and computation-intensive services are growing rapidly, and MEC-enabled base stations can meet the demand for low-delay and highly reliable power services. Because the capacity of base stations is limited, most existing schemes use queuing or retransmission to reduce the pressure of base stations. However, the quality of service (QoS) will deteriorate due to queuing and retransmission. In this paper, we study the cooperation between MEC base stations to improve the computing service capability by offloading computing tasks to other MEC base stations. An optimization strategy based on the Q-learning algorithm is proposed to reduce the total delay of power user equipment (UE) within the coverage of the base station. We consider the limitations of terminal battery capacity and the computing capacity of base stations. The simulation results show that the strategy effectively reduces the system delay and achieves the load balancing of the base station.