Research On Wireless Resource Management Technology For VR Video Service

The deployment of Virtual Reality(VR)applications in wireless network environment can provide users with immersive environment without behavioral restrictions,which has attracted considerable attentions in recent years.Different from traditional video services,VR video requires ultra high bandwidth and low latency,which brings more and more pressure to wireless networks.Traditional capacity expansion methods such as increasing the number of base stations and bandwidth cannot effectively cope with this challenge.Network resource optimization based on VR service characteristics is an important technical approach to solve the above problems.Considering that a large number of users often request the same content at the same time in VR live casting scenes,the adoption of multicast technology can improve the spectral efficiency.Thus,it is of great significance to study how to effectively combine IT with VR video transmission to improve system performance.At the same time,for VR ondemand services,the caching capability at the network edge can reduce service delay and improve user experience.Therefore,the research on the VR video transmission strategy based on edge caching is also a key approach to optimize the system performance.This thesis carries out optimization designs of VR video transmission respectively for VR live multicast and on-demand scenarios,combined with different requirements and characteristics.First of all,a multicast transmission system based on MPEG-I video format is considered,and a joint user grouping and bandwidth allocation algorithm is proposed to improve the user terminal experience of VR.Specifically,the scheme includes optimal bandwidth allocation and heuristic user grouping based on logarithm utility.A multicast double-layer packet scheme is proposed for user grouping.According to the grouping results and theoretical analysis,the optimal bandwidth ratio of each group is proportional to the number of users in the group.In addition,considering the special case that a single base station serves a large number of VR users with high SNR at the same time,a simple and effective grouping algorithm is proposed,which greatly reduces the channel feedback cost.Simulation results show that the proposed scheme can effectively improve user utility under the premise of the same bandwidth.Compared with the ergodic search method,the proposed method achieves similar performance with low computational complexity under high SNR conditions.Secondly,combined with the wireless edge caching technology,a joint cell access and version selection strategy based on field of view(FOV)is designed for the wireless VR video-on-demand scenario to improve the total utility of the network.Firstly,the quality of different video versions,wireless cache and other factors are considered comprehensively to maximize the sum utility of the network.An integer optimization problem is formulated and derived and transformed into a multiple choice knapsack(MCK)problem and a matching problem,and a distributed solution is carried out.Based on Knapsack algorithm and switching matching theory,a joint version selection and user access algorithm for mobile wireless VR on-demand service is proposed.The simulation results show that compared with the baseline,the proposed algorithm improves the total utility of the system,increases the number of users of network service,and finally balances the traffic load effectively.