一种移动边缘云计算任务卸载算法

A TASK OFFLOADING ALGORITHM IN MOBILE EDGE CLOUD COMPUTING

移动边缘云计算环境中,多个移动用户竞争无线通信资源时,多信道无线干扰会影响移动用户的任务卸载能效。移动用户如何作出任务卸载决策并选择合适的无线信道进行数据上传是必须解决的难题。针对该问题,将边缘云环境中的任务卸载问题形式化为信道干扰和延时约束下的0-1非线性整数规划问题,并设计一种基于能效的任务卸载算法。为求解该问题,算法设计了移动用户分类和优先级确定机制,并在此基础上,利用拍卖理论求解了任务卸载决策,并确定了最优的任务卸载信道。实验仿真结果表明,与对比算法相比,该算法在移动设备的平均能耗、任务执行延时以及移动设备吞吐量三个指标上均表现出更好的性能优势。

In mobile edge cloud environment,when multiple mobile users compete for communication resources,the multichannel wireless interference will affect the energy efficiency of task offloading of mobile users.How to make the task offloading decision for each mobile user and select the suitable wireless channel for data uploading become a problem that must be solved.Aiming at the problem,we formulate the problem of the task offloading in mobile edge cloud environment as a 0-1 nonlinear integer programming problem under the constraints of channel interference and the delay,and design the task offloading algorithm based on energy efficiency.To solve the problem,we devised the classification and priority determination mechanism for the mobile users.And on the basis of that,the auction theory was used to solve the task offloading decision and determine the optimal channel for task offloading.Experimental simulation results show that compared with the comparison algorithms,our algorithm shows better performance on three metrics,which are the average energy consumption of mobile devices,the task execution delay and the throughput of mobile devices.