边缘计算:万物互联时代新型计算模型

发布时间：2018-05-14 22:38

本文选题：边缘计算 + 云计算　；参考：《计算机研究与发展》2017年05期

【摘要】：随着物联网的快速发展和4G/5G无线网络的普及,万物互联的时代已经到来,网络边缘设备数量的迅速增加,使得该类设备所产生的数据已达到泽字节(ZB)级别.以云计算模型为核心的集中式大数据处理时代,其关键技术已经不能高效处理边缘设备所产生的数据,主要表现在:1)线性增长的集中式云计算能力无法匹配爆炸式增长的海量边缘数据;2)从网络边缘设备传输海量数据到云中心致使网络传输带宽的负载量急剧增加,造成较长的网络延迟;3)网络边缘数据涉及个人隐私,使得隐私安全问题变得尤为突出;4)有限电能的网络边缘设备传输数据到云中心消耗较大电能.为此,以边缘计算模型为核心的面向网络边缘设备所产生海量数据计算的边缘式大数据处理应运而生,其与现有以云计算模型为核心的集中式大数据处理相结合,即二者相辅相成,应用于云中心和网络边缘端的大数据处理,较好地解决了万物互联时代大数据处理中所存在的上述问题.边缘计算中的"边缘"是个相对的概念,指从数据源到云计算中心数据路径之间的任意计算资源和网络资源.边缘计算的基本理念是将计算任务在接近数据源的计算资源上运行.首先系统地介绍边缘计算的概念和原理;其次,通过现有研究工作为案例(即云计算任务迁移、视频分析、智能家居、智慧城市、智能交通以及协同边缘),实例化边缘计算的概念;最后,提出边缘计算领域所存在的挑战.该文希望能让学界和产业界了解和关注边缘计算,并能够启发更多的学者开展边缘式大数据处理时代边缘计算模型的研究.
[Abstract]:With the rapid development of the Internet of things and the popularity of 4G/5G wireless network, the age of interconnection of all things has arrived, and the number of network edge devices has increased rapidly, which makes the data generated by this kind of devices have reached the level of ZB. In the era of centralized big data processing based on cloud computing model, its key technology can no longer efficiently handle the data generated by edge devices. The linear growth of centralized cloud computing capabilities can't match the explosive growth of massive edge data.) the transmission of massive data from network edge devices to the cloud center results in a sharp increase in the load of network bandwidth. The network edge data involves personal privacy, which makes the privacy security problem more prominent. 4) the network edge devices with limited power consumption consume a large amount of power to transmit data to the cloud center. Therefore, the edge big data processing based on the edge computing model, which is oriented to the massive data computing generated by the network edge devices, is combined with the existing centralized big data processing based on the cloud computing model. That is to say, they complement each other and are applied to the big data processing in the cloud center and the edge of the network, which solves the above problems in the big data processing in the age of interconnection of all things. "Edge" in edge computing is a relative concept, which refers to arbitrary computing resources and network resources from data sources to cloud computing center data paths. The basic idea of edge computing is to run computing tasks on computing resources close to the data source. First, the concept and principle of edge computing are introduced systematically. Secondly, through the existing research work as a case (i.e. cloud computing task migration, video analysis, smart home, smart city, The concept of intelligent traffic and collaborative edge computing is instantiated. Finally, the challenges in the field of edge computing are presented. This paper hopes to make the academic and industry understand and pay attention to edge computing, and can inspire more scholars to study edge computing model in the era of edge big data processing.
【作者单位】：韦恩州立大学计算机科学系;安徽大学计算机科学与技术学院;
【基金】：国家自然科学基金面上项目(61572001) 安徽大学2016年博士科研启动经费项目(J01003214)~~
【分类号】：TP3

【相似文献】