Research On Panoramic Video Streaming Over DASH

Video services have become the most important form of business for mobile networks today.With the emergence and development of Virtual Reality(VR),the panoramic video service has evolved into an independent online application.As a new consumer growth point in the mobile video market,the panoramic video is expected to become a typical business of 5G networks.However,the panoramic video transmission under wireless networks still meets two severe challenges.First,the panoramic video has a huge amount of data which makes the wireless network fail to support its real-time transmission.Second,the time-varying problem of the wireless network will still have a huge impact on the quality of service.Based on the research of panoramic video,projection,video coding,and streaming technology,a panoramic video adaptive transmission scheme based on Truncated Square Pyramid Projection(TSP)and HTTP Dynamic Adaptive Streaming over HTTP(DASH)is proposed in this paper.The innovation of this scheme lies in its adaptive characteristics,which are embodied in TSP-based adaptive compression and DASH-based adaptive bitrate adjustment.The former solves the problem of the inadequacy of network bandwidth by reducing the requirement for transmission;the latter solves the problem of re-buffering caused by bandwidth fluctuation in the wireless network and improves the Quality of Experience(QoE).The main work of this paper includes four aspects.First,the panoramic video adaptive transmission scheme mentioned above is proposed in this paper.Second,the feasibility and adaptive compression characteristics of TSP in this scheme are analyzed.The TSP is compared with the traditional proj ection from the perspective of geometric distortion and bandwidth saving.Third,aiming at the problem of instantaneous bitrate fluctuation of the panoramic video,a server-assisted instantaneous bitrate prediction algorithm is proposed.Then,based on that,bandwidth estimation,bitrate selection,and buffer control strategy are designed to form a DASH adaptation algorithm;Fourth,the pedestrian and vehicle simulation scenarios under LTE network are built with ns-3.The simulation model of a viewpoint-based panoramic video transmission system is also designed and implemented.Then the simulation and verification of DASH adaptation algorithm are completed.