Bipartite Graph based Edge Server Placement Algorithm in Mobile Edge Computing
Ke Yin · 2021 International Conference on Electronic Information Engineering and Computer Science (EIECS) · 2021
In the context of smart application, a large amount of data is generated due to enormous mobile users' access, causing increased service latency with the traditional cloud computing framework. To reduce network and service latency, edge computing has been proposed to offload part of the workload generated by mobile users to nearby edge servers with sufficient resources. Although mobile edge computing has been significantly studied, there is little attention given to the placement of edge servers in smart applications to optimize the service performance. In this paper, we investigate the edge server placement problem in edge computing based on bipartite graph. The edge server placement problem is firstly formulated as a graph optimization problem. Then, we utilize a bipartite graph model to find the optimal solution. Experimental results based on open dataset and simulated datasets show that, with the bipartite graph-based edge server placement, the proposed approach outperforms counterparts in terms of service access delay and system workload balancing.