无人机边缘计算:架构、多址接入与计算卸载

Unmanned aerial vehicle-enabled edge computing:architecture,multiple access and computation offloading

无人机(unmanned aerial vehicle,UAV)边缘计算技术将UAV平台与移动边缘计算技术相结合,充分利用UAV的灵活性和机动性,为用户设备提供及时有效的计算服务。从UAV边缘计算的网络架构入手,提出基于网络功能虚拟化和软件定义网络的技术架构。针对UAV边缘计算的关键技术,总结对比UAV边缘计算中不同的多址接入方案,并从不同的优化目标出发,对基于经典非凸优化、博弈论以及人工智能方法的计算卸载策略进行总结和分析。最后,探讨和展望未来的研究方向。

Unmanned aerial vehicle(UAV)-enabled edge computing combines mobile edge computing technology with UAV platform.It can provide timely and effective computing services for wireless devices by taking advantages of the flexible deployment and mobility of UAV.Starting with the network architecture of UAV-enabled edge computing,the technical architecture based on network function virtualization and software-defined networks is proposed.For the key technologies of UAV-enabled edge computing,different multiple access schemes in UAV-enabled edge computing network are compared.Furthermore,according to different optimization objectives,the computation offloading strategies based on classical non-convex optimization,game theory,and artificial intelligence are summarized and analyzed.Finally,the future research direction is discussed and prospected.