随着5G网络的商用普及，高质量的虚拟现实视频服务逐渐成为可能。为了达到传统高清视频的质量，需要传输16K/24K分辨率的虚拟现实视频，这对网络、终端和服务器都带来巨大的带宽、计算压力。针对这些挑战，本文介绍了虚拟现实视频传输的质量评价和面临的挑战，并详细介绍了两种自适应传输关键技术：（1）多层FOV传输；（2）多用户协作传输。重点解决自适应传输中的卡顿、黑场、多用户竞争等问题，提高虚拟现实视频传输的用户体验质量(Quality-of-Experience)。

With the deployment of commercial 5G network, Virtual Realty Video (VR Video) are coming to our life. To achieve the equal quality as HDTV, VR video needs 16K/24K resolutions which introduce significant overhead on network, server and terminal. This paper presents the quality metrics and challenges of viewport-adaptive VR video streaming. Two viewport-adaptive streaming methods are proposed: 1, Multi-layer FoV-adaptive streaming. 2, Multi-user Cooperative Streaming with reinforcement learning. They address the quality impairment of stalling, black hole and mutli-user competition and improve the quality-of-experience of VR video streaming.