An efficient module deployment algorithm in edge computing

Abstract

Edge computing refers to data storage and processing at the edge of a network instead of cloud data centers. In edge computing, user data can be preprocessed to reduce the network traffic and load of the cloud system. Module deployment scheme is vital in edge computing since it has a significant effect on the performance and the resource utilization of the edge computing system. Our objective is to maximize satisfied user requests under the constraints of limited communication bandwidth and storage capacity of devices where the modules are deployed. Since the optimization problem is NP-hard, we propose a heuristic algorithm based on relaxation and rounding techniques to solve this problem. The simulation results show that our algorithm satisfies more user requests than previous work.