面向物联网的移动边缘计算任务卸载方法研究

Research on Task Off Loading Method of MEC in Internet of Things

为了有效地解决移动设备计算能力和存储能力不足等问题,以物联网的移动边缘计算(mobile edge computing,MEC)为出发点,对边缘计算卸载算法相关研究现状与成果进行了综合分析,深入剖析MEC卸载机制,从任务卸载算法的输入参数、输出参数、卸载算法、性能指标四个角度详细分析MEC卸载的研究现状,并结合卸载原理分析MEC的应用场景,指出了边缘计算卸载的未来研究方向:融合深度学习和强化学习的MEC卸载、面向5G网络的边缘计算卸载算法、边缘智能在分布式环境下的并行化.

In order to effectively solve such problems as lack of mobile computing power and storage capacity,the calculation is based on the mobile edge of the Internet of things(mobile edge computing,MEC)as a starting point,on the edge of computing offloading technology related research present situation and the results are analyzed,in-depth MEC offloading mechanism,and from the task of offloading algorithm input parameters,output parameters,offloading algorithm,the performance index of four point of view.Combined with the offloading principle,the application scenarios of MEC are analyzed,and the future research directions of edge computing offloading are pointed out:MEC offloading integrating deep learning and reinforcement learning,edge computing offloading technology oriented to 5G network,parallel of edge intelligence in distributed environment.