基于多目标优化的移动边缘计算任务卸载方法

Mobile edge computing task offloading method based on multi⁃objective optimization

为了缩短任务卸载时延和降低能耗,基于多目标优化设计一种移动边缘计算任务卸载方法。首先,建立一个多目标优化模型,以最小时延和最小能耗为目标函数,同时考虑任务卸载计算资源分配约束、总能耗低于设备剩余电量约束、任务卸载传输功率约束、任务卸载时延约束等约束条件;考虑了移动设备的移动性,构建节点切换约束条件;为了提高求解效率,改进了遗传算法中的选择、交叉和变异操作,并利用改进后的遗传算法求解了多目标优化模型,得到了最优的移动边缘计算任务卸载策略。实验结果表明,该方法能够有效地完成任务卸载,并且能够减少任务卸载时延和降低任务卸载能耗。此外,该方法在不同剩余电量下都能够获得较好的任务卸载效果,从而提高了用户的体验。

In order to shorten the task offloading delay and reduce energy consumption,a mobile edge computing(MEC)task offloading method is designed based on multi⁃objective optimization.A multi⁃objective optimization model is established with minimum delay and minimum energy consumption as objective functions,while considering constraints such as task offloading computing resource allocation,total energy consumption being lower than device remaining power,task offloading transmission power and task offloading delay.In addition,the mobility of mobile devices is considered,and node switching constraints are constructed.In order to improve the solving efficiency,the selection,crossover and mutation operations in the genetic algorithm are improved,and the multi⁃objective optimization model is solved with the improved genetic algorithm,and the optimal task offloading strategy of mobile edge computing is obtained.The experimental results show that the proposed method can complete task offloading effectively,shorten the task offloading delay and reduce the task offloading energy consumption.In addition,it can achieve good task offloading performance under different remaining battery levels,so it can improve the user experience.