Research on Container Migration Mechanism of Power Edge Computing on Load Balancing
Kai Li, Chunlei Chang, Kai Yun, Jianye Zhang
Abstract
To resolve some problems of traditional cloud computing such as resource bottleneck, edge computing is proposed, which is very useful in the field of power Internet of Things. Because the equipment is unevenly distributed geographically and the demand for computing resource and storage resource is different, the load rate on different edge nodes has great difference. The development of container technology solves the problem of load imbalanced. Because of the features of lightweight and portability, the containers can realize the dynamic task assignment according to node load rate, and then realize the load balance of edge nodes. Therefore, a decision mechanism based on container migration and a migration algorithm is proposed in this paper. The optimized migration of containers can be realized by improving ant colony algorithm, setting up local pheromone volatilization model and global pheromone updating model. The simulation results show that PCM has a better performance on improving load balancing and reducing migration overhead.