A Potential Game Based Load Balancing Approach for Mobile Edge Computing Enabled Intelligent Rail Construction System

Empowered by mobile edge computing (MEC) technology, the intelligent railway construction system can further improve railway construction efficiency and ensure worker safety. By offloading complex computation tasks from smart devices (e.g., smart helmets and construction machinery) to a physically adjacent MEC server, the railway construction system can provide highly reliable and low-latency service for the construction machinery and workers. Due to the massive number of computing tasks from smart devices and machinery, MEC servers cannot always handle all the incoming computation tasks on time when they operate independently. They often need to cooperate through peer-offloading, and load balancing among multiple MEC servers becomes a critical challenge. In this paper, we design an MEC-based intelligent railway construction system. Several intelligent construction scenarios are designed using an MEC-based computing service. With the aim to minimize the average task computing time, we propose a potential game model to optimize the load balance among MEC servers. Multiple computation tasks are scheduled among the servers, and a learning algorithm is used to obtain the Nash equilibrium where the best response of each server is achieved. Extensive experimental results demonstrate that the proposed model can improve the performance of the intelligent railway construction system.