Task Offloading and Power Assignment Optimization for Energy-Constrained Mobile Edge Computing

Mobile edge computing (MEC) can deliver low-latency computing services by deploying edge servers that are close to mobile devices. This paper presents a task offloading and power assignment optimization algorithm for minimizing task completion time under an energy constraint upon the mobile device. We develop a grey wolf optimizer (GWO)-based metaheuristic algorithm to determine the offloading order and power assignment for task offloading. The proposed algorithm employs a position-based mapping scheme for converting each grey wolf into a high-quality offloading solution represented by a task sequence. For the ordered tasks in the converted solution, we develop a heuristic strategy to assign their transmission powers such that the fitness value of the corresponding grey wolf can be evaluated. We further introduce a selection mechanism for updating the wolf population, in which grey wolves with high fitness values have high probabilities of being selected in the next generation of grey wolves. Evaluation results demonstrate that the new mapping scheme and selection mechanism are beneficial for improving task offloading performance in terms of significant makespan reductions.