Virtual Reality Video Oriented Cross-layer Transmission Optimization Over Wireless Networks

With the development of mobile terminals and the increasing enrichment of online video content,the demand for streaming video service applications is increasing,and it is becoming the main type of traffic in wireless networks.The rapid development of virtual reality(VR)wearable devices makes VR streaming video service applications quickly enter people's sights and lives,and is widely used in video live broadcast,immersive games,online exhibition halls,panoramic street scenes and other fields.However,with the widespread application of streaming video services and the explosive development of content,the contradiction between business needs and wireless network transmission capabilities has become increasingly prominent.As a new multimedia service,the transmission of VR streaming video services faces two major challenges:great bandwidth requirements and strict delay requirements.The existing transmission network is difficult to meet the business needs through direct transmission,otherwise it will sacrifice the picture quality,or there will be serious picture stagnation and scene switching delay,causing the viewer's dizziness.However,as a viewer,it is hoped that the picture quality presented will be vivid and delicate,and the video information transmitted will be rich and instant.In addition,when multiple Dynamic Adaptive Streaming over Hyper Text Transfer Protocol(DASH)streams request downloading together,competition between them will lead to bottleneck bandwidth,reduce bandwidth utilization,and sacrifice the performance of the DASH service,which affects the user's service experience.How to achieve a balance between bandwidth utilization and user viewing experience,especially to solve the competition coordination between multiple DASH streams,is also one of the research focuses of this topic.In response to the above challenges,this paper establishes an auxiliary communication mechanism between the service and the network based on DASH,from the perspective of service adaptation for network and network adaptation for service.That is,cross-layer optimization,design and implementation of two wireless network cross-layer optimization schemes for VR video transmission,based on the traditional Open System Interconnection Reference Model(OSI),to achieve optimization.Firstly,taking the naked-eye three dimension(3D)video application as an example,QoE-driven and Network-assisted Naked-eye 3D Adaptive Video Streaming over Wireless Network(QNANA)is proposed from the perspective of service adaptation for network.This transmission optimization solution can not only reduce the service requirements of the network bearer capacity,improve resource utilization,but also ensure the quality of user viewing experience.The experimental results show that QNANA can significantly improve the bandwidth prediction accuracy,thereby improving resource utilization and significantly improving the QoE.Secondly,taking the panoramic video application as an example,from the perspective of network adaptation for service,the Viewpoint-Feedback based Multi-User Non-Cooperative Resource Scheduling Algorithm(VFNCR)optimization scheme is proposed.VFNCR establishes a feedback mechanism for business information to improve the performance of the business while ensuring efficiency and fairness in the process of resource allocation.By establishing a forward feedback loop,the VFNCR algorithm maintains the health status of the cache queue on the service side,avoids delays and reduces the delay,improves resource utilization,and can balance system throughput and resource allocation fairness.Experimental results show that performance has improved.