Numerical Simulation Design of Multiple Users Offloading Using Improved Optimization Approach for Edge Computing

Abstract

It’s no secret that the IoT (Internet of Things) has exploded in popularity over the past decade. IoT devices with limited resources have a lot of trouble kee** up with the growing number of latency-sensitive and computationally intensive IoT applications. By allowing devices with constrained resources to outsource their work to edge servers, edge computing looks to be a viable technique for expanding the computational capacity of IoT systems. Most of the existing literature on task offloading overlooks the interdependencies between tasks and subtasks, despite the fact that they provide a significant difficulty and may have a considerable influence on offloading decisions. Furthermore, the current research commonly considers offloading activities to specific edge servers, which may result in underutilization of edge resources in very busy edge networks. In this research, we look at the problem of offloading tasks in dense edge networks while kee** dependencies in mind. To achieve full parallelism between edge servers and IoT devices, we measure task dependency using directed acyclic graphs (DAGs). In order to reduce both task delay and energy consumption, task offloading is frequently given as a joint optimization issue. We prove that this is an NP-hard issue and present a heuristic approach to guaranteeing subtask dependency while improving task efficiency. The suggested strategy is shown to be effective in reducing task latency in simulations of highly dense edge networks.