Delay-aware Scheduling Scheme for Ubiquitous IoT Applications in Edge Computing
Ali Abdulnaser Alnoman · 2021
Edge computing emerged as a promising solution for the ubiquity of IoT applications by providing cloud-like services in close proximity to users thus reducing delay and relieving cloud servers. In this paper, a task scheduling scheme in edge computing is proposed to accommodate the unprecedented amounts of IoT tasks taking into account the computing requirements of these tasks; in particular, the task completion deadline. To this end, a 0-1 knapsack problem is formulated to optimally allocate the available edge resources considering the task size and completion deadline. The problem is solved using dynamic programming wherein tasks with shorter deadlines are given higher values to prioritize delay-sensitive applications. The solution is compared against a baseline approach which is the first-in first-out scheduling scheme. Results show the effectiveness of the proposed scheme in increasing the overall number of accommodated tasks as well as the number of accommodated delay-sensitive tasks.