Task Offloading Based on Utility Optimization in Collaborative Edge Computing Systems

In collaborative edge computing systems, offloading computation tasks from users to edge nodes can effectively address the issues of insufficient terminal computing capabilities and high latency, high overhead of cloud computing processing. System overhead and utility metrics are critical factors that impact operational efficiency and inform task offloading decisions. This paper proposes a new utility evaluation model and, by considering task completion time constraints and energy consumption overhead, establishes an overhead-utility balanced optimization method to address task offloading problem and enhance system operational efficiency of collaborative systems. We propose a heuristic solution based on simulated annealing and particle swarm optimization to solve the overhead-utility balanced optimization problem effectively. Through experimental simulation testing, our proposed heuristic solution can efficiently achieve the balance between low system overhead and high utility in making task offloading decisions, fully utilize system resources, and enhance operational efficiency.