Joint Optimization of Service Caching and Task Scheduling in Energy-Efficient Mobile Edge Computing

This paper studies an energy-efficient task scheduling problem that takes into account the cooperation among service caching-enabled mobile edge computing (MEC) servers. We consider a MEC system consisting of multiple sub-networks of MEC nodes and formulate the studied problem as a constrained optimization model that makes decisions on energy-efficient assignment of task requests. To cope with the formulated optimization problem, we propose a metaheuristic approach that is built upon the searching mechanism of the squirrel search algorithm (SSA) but involves improved operators and name the metaheuristic approach ISSA. By using an offloading sequence (i.e., a permutation of all computation tasks) to represent the scheduling solution, ISSA uses a cooperative task assignment heuristic for evaluating the fitness value of each candidate solution. This heuristic is responsible for dispatching each task in the sequence to a certain MEC node within the sub-network or outsourcing the task to the cloud in case of insufficient resources, such that the total energy consumption can be calculated. In addition, when updating the squirrel positions to facilitate the metaheuristic search, we employ a new updating rule to enhance ISSA’s capability in exploring the solution space. We perform extensive simulations to demonstrate the advantage of our proposed ISSA over other metaheuristic approaches in reducing the total energy consumption.