Design of a service caching and task offloading mechanism in smart grid edge network

Abstract

Smart Grid Communication System (SGCS) needs to meet the QoS requirements of various applications in terms of latency, bandwidth and reliability. However, cloud computing owns significant latency and cannot meet the need of real-time applications, edge computing is gradually applied in SGCS. In this paper, LBPC (edge-based load-balancing algorithm based on popularity and centrality) is proposed based on edge network and achieves the deployment of computing units at the edge nodes. In the aspect of service caching, LBPC measures the popularity of requests and the centrality of nodes, and it also takes the latency needs and the cache cost into consideration. In the aspect of task offloading, LBPC calculates the cost at different neighbor nodes and chooses the best one to finish the calculation when the current node is in high-load condition. Compared with some related works, experimental results show that LBPC can effectively reduce communication latency and balance the network load.