Abstract:

The rapid development of science and technology has made the concept of the Internet of everything no longer just appear in people??s mind. With the rapid growth number of smart devices accessing to the wireless network, the amount of edge data has reached the ZB level, which puts tremendous pressure on the core network bandwidth; at the same time, the emergence of many emerging applications such as driverless, location recognition, augmented reality, virtual reality, etc. puts higher demands on network delay, jitter and data security. Traditional cloud computing can no longer meet the requirements, so edge computing (EC) came into being. Edge computing provides lightweight cloud computing and caching capabilities at the edge of networks. In this survey, an exhaustive review on the state-of-the-art research efforts on edge computing is made. First, the concepts of edge computing and cloud computing are summarized, the advantages of edge computing are compared and analyzed, the inevitability and trend of the development of edge computing are pointed out. Then a comprehensive overview of typical edge computing architectures and platforms is presented, and typical application cases in edge computing such as network performance optimization, video caching, shopping cart view refresh and Web video live broadcast are discussed. Finally, from the aspects of edge computing service management, application mobility management, computing resource management and data management, this paper looks forward to the open research challenges and future development trends of edge computing, hoping to bring inspiration to researchers engaged in edge computing.

Key words: edge computing, mobile edge computing (MEC) architecture, computation offloading, intelligent nodes, edge caching

摘要：

科学技术的快速发展，使万物互联设想不再仅仅停留在人们的概念中。随着接入无线网的智能设备数量的快速增长，边缘数据量已达到ZB级别，给核心网络带宽造成巨大压力；与此同时，无人驾驶、位置识别、增强现实、虚拟现实等众多新兴应用的出现对网络延迟、抖动、数据安全等提出了更高的要求。传统云计算在以上方面表现乏力，于是边缘计算（EC）应运而生。边缘计算能够在网络的边缘提供轻量级的云计算和存储能力。对边缘计算的最新研究成果和应用进行了详尽的回顾。首先综述边缘计算和云计算的概念并对比分析边缘计算的优势，指出边缘计算发展的必然性和时代趋势；然后针对典型边缘计算架构和平台进行了全面的综述，并讨论了网络性能优化、视频缓存、购物车视图刷新和网络视频直播等边缘计算中的典型应用案例。最后，从边缘计算服务管理、应用移动性管理、计算资源管理、数据管理等四方面，展望了边缘计算的开放式研究挑战和未来的发展趋势，希望能给从事边缘计算的科研工作者带来启发。

关键词: 边缘计算, 移动边缘计算（MEC）架构, 计算迁移, 智能节点, 边缘缓存