面向自动驾驶的边缘计算技术研究综述  

Survey on edge computing technology for autonomous driving

作　　者：吕品[1,2,3] 许嘉 李陶深[1,3] 徐文彪 LYU Pin;XU Jia;LI Taoshen;XU Wenbiao(School of Computer,Electronics and Information,Guangxi University,Nanning 530004,China;Guangxi Key Laboratory of Multimedia Communications and Network Technology,Nanning 530004,China;Guangxi Colleges and University Key Laboratory of Parallel and Distributed Computing,Nanning 530004,China)

机构地区：[1]广西大学计算机与电子信息学院,广西南宁530004 [2]广西多媒体通信与网络技术重点实验室,广西南宁530004 [3]广西高校并行分布式计算技术重点实验室,广西南宁530004

出　　处：《通信学报》2021年第3期190-208,共19页Journal on Communications

基　　金：国家自然科学基金资助项目(No.62062008,No.61771312,No.62062006);广西自然科学基金资助项目(No.2018JJA170194,No.2018JJA170028,No.2019JJA170045);广西八桂学者专项经费资助项目。

摘　　要：边缘计算在自动驾驶的环境感知和数据处理方面有着极其重要的应用。自动驾驶汽车可以通过从边缘节点获得环境信息来扩大自身的感知范围,也可以向边缘节点卸载计算任务以解决计算资源不足的问题。相比于云计算,边缘计算避免了长距离数据传输所导致的高时延,能给自动驾驶车辆提供更快速的响应,并且降低了主干网络的负载。基于此,首先介绍了基于边缘计算的自动驾驶汽车协同感知和任务卸载技术及相关挑战性问题,然后对协同感知和任务卸载技术的研究现状进行了分析总结,最后讨论了该领域有待进一步研究的问题。Edge computing plays an extremely important role in the environment perception and data processing of autonomous driving.Autonomous driving vehicles can expand their perception scope by obtaining environmental information from edge nodes,and can also deal with the problem of insufficient computing resources by offloading tasks to edge nodes.Compared with cloud computing,edge computing avoids high latency caused by long-distance data transmission,and provides autonomous driving vehicles with faster responses,and relieves the traffic load of the backbone network.Firstly,the edge computing-based cooperative perception and task offloading technologies for autonomous vehicles were introduced firstly,and related challenging issues were also proposed.Then the state-of-the-art of cooperative perception and task offloading technologies were analyzed and summarized.Finally,the problems need to be further studied in this field were discussed.

关 键 词：自动驾驶 边缘计算 协同感知 任务卸载 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]