Exploring Delay Reduction on Edge Computing Architectures from a Heuristic Approach.

Edge computing is a new promising paradigm helps users to execute their tasks on edge network which is closer to them rather than cloud. It can reduce application response time especially for those are critical to time such as healthcare applications, real-time apps, game playing, or traffic systems. Edge User Allocation (EUA) problem is responsible for allocate user application into edge servers on the edge network as app vendors’ needing. In this paper, we propose a heuristic called Nearest Edge Server with Highest Capacity (NESHC) to solve the EUA problem. We use a real-word dataset in our extensive experiments. The results show that NESHC can reduce elapsed CPU time and outperform baseline approach (Optimal) and two state-of-the-art approaches (ICSOC19 and TPDS20). The reduction of delay causes increasing of user allocated into edge servers and leveraging the overall utilization of edge network system.