Modeling And Optimization Of Scalable Video Multicast In LTE Systems

In recent years,with the rapid development of wireless communication technolo-gies,the data-rate of mobile networks has been greatly improved,and mobile video services are also growing rapidly.The bandwidth consumption of video traffic is large,however,the repeatability of the transmission is also high.It is necessary to take mea-sures to realize the effective use of wireless resources to meet the high bandwidth and low latency requirements of the transmission of mobile video services.As the lead-ing technology of next-generation mobile communications,LTE has been rapidly com-mercialized globally.Studying how to improve video transmission efficiency in LTE networks is of great significance to operators,service providers,and users.In fact,for some scenarios like mobile TV,live broadcast,to transmit the same video data to a large number of users,multicast technology can effectively improve video transmis-sion efficiency.LTE's eMBMS extension provides support for multicast and broadcast services.However,affected by cell users,the resource allocation policy for multicast in LTE/eMBMS systems is usually inefficient.This thesis combines Scalable Video Coding(SVC)and channel feedback informa-tion to propose an efficient layered multicast scheme.Assigning a low rate Modulation and Coding Scheme(MCS)to the base layer allows most users to receive the base layer and get the basic video quality.Assigning a high rate MCS to the enhancement layer enables users with good channel conditions to receive more layers and get better video quality.The main work of this thesis is summarized as follows:The LTE/eMBMS system layered multicast architecture is introduced in detail.The new features added by the server,terminal,and base station to support layered multicast services are described.The server and the terminal mainly need to add SVC codec support,while the base station needs to redesign the resource allocation strategy of layered multicast service.Then we establish a mathematical model for the most critical part of layered mul-ticast,ie,allocation of downlink resources.The existing research methods for solving the problem of resource allocation of layered multicast are basically based on determin-istic methods and do not take into account the stochastic of the wireless channel.We models the problem as a stochastic optimization model and sloves the model with Lya-punov drift and optimization theory.For the sake of complexity,the resource allocation algorithm of single group scenario is obtained by minimizing drift plus penalty and a two-stage resource allocation algorithm is designed for multiple groups scenario.The first stage pre-allocates RBs for each group.The second stage invokes the resource al-location algorithm of single group scenario.The overall complexity is low.In addition,in order to improve the spectrum efficiency of the system,some optimizations are made for the MCS and RBs allocation scheme of the algorithm.A simulation system was implemented on Matlab,and modules such as link adap-tation,downlink scheduling module were implemented.Experiments was designed and performed for different numbers of multicast groups and different bandwidth configu-rations,and compared with the comparison algorithm in video quality and bandwidth occupancy.The experimental results show that the proposed algorithm can provide higher video quality and better service for users in the same bandwidth occupation.In addition,a parameter sensitivity experiment was designed.The experimental results show that tradeoff can be made between the bandwidth occupation and video quality by adjust the parameter.