Research On Multimedia Service In High-Speed Mobile Networks

In recent years,the rapid development and wide application of advanced transportation technologies such as high-speed train have boosted great demands for high-quality Internet services in High-Speed Mobile(HSM)Networks.More and more mobile users are hoping to use Internet services with high qualities to enjoy their journey.However,the high mobility of HSM networks greatly reduces the service quality.For example,the high－speed train’s velocity is about 300km/h-350km/h.Such a high speed will significantly enhance the negative influence of the Doppler effect on the communication bandwidth.Moreover,the handover frequency will also be increased which will further decreases communication quality.Therefore,it is still challenging to provide Internet services with high-quality in HSM networks.Meanwhile,among the huge number of network service types,multimedia services occupy the mainstream position.According to Cisco’s report,80%of current network traffic comes from multimedia services.However,the high bandwidth and low latency characteristics of multimedia services make it difficult for users to obtain a good quality of service experience in HSM networks.Although numerous dynamic adaptive multimedia technologies have been proposed,the highly dynamic network state in HSM networks has caused great difficulty for dynamic adaptive multimedia technologies to achieve real-time optimization of multimedia services.Based on the above problems,this thesis takes the typical HSM network scenario,i.e.High-Speed Railway(HSR)network,as an example,and conducts in-depth research in wireless data transmission and multimedia service optimization.The HSR network researched in this paper is the data interaction network used to provide network services for high-speed train passengers.In the aspect of wireless data transmission,firstly,for the data loss problem on the wireless link,a packet loss-aware adaptive scalable transmission method is proposed.Then,for the multi-base station access model and data scheduling problem,a multi-base station data dissemination method is proposed.In the aspect of multimedia service optimization,research is carried out from the user side and the service side respectively to comprehensively improve the performance of the multimedia service in HSR network.On the user side,considering the quality of user service and the characteristics of multimedia service technology,a stochastic optimization mechanism for video service based on dynamic adaptive technology is proposed.On the service side,a new business model in which service providers and network operators cooperate is analyzed.Then with multimedia service profit as the main optimization goal,a stochastic profit optimization mechanism for multimedia services is proposed.Based on the above four aspects of research,this thesis has formed a set of HSR network service performance optimization research system.The research works are detailed as follows:(1)Through the Packet Loss Rate(PLR)test on the Beijing-Shanghai HSR line,it is found that high packet loss is the main problem of HSR network.Further analysis of HSR network architecture indicates that the wireless link between the train and base station is the main cause of high packet loss problems.Then based on the automatic repeat-request protocol in data link layer,a mathematical model is established and a nonlinear programming problem is constructed with the minimum data transmission time as the optimization objective.Through the analysis of the problem,it is found that when the PLR is higher than a certain threshold,redundant replication of the data packets can effectively improve the network transmission efficiency.Based on such finding,the Loss-Aware Adaptive Scalable Transmission mechanism is constructed.Finally,the effectiveness of the proposed transmission mechanism is verified by simulation experiments.(2)It is found that the current research on data dissemination and transmission of HSR network is mainly carried out under the single-base station model,that is,network service data can only be disseminated to selected base station.Although this method can reduce the difficulty of network model analysis to some extent,it ignores the problem of base station node selection when multiple base stations can establish wireless transmission links with users.Therefore,the network model of multiple base stations is first established which can be used for wireless data transmission.Then,considering the wireless link data transmission power and data loss probability,a multi-base station data dissemination algorithm is proposed and verified by simulated experiments.(3)The newly proposed Dynamic Adaptive Streaming over HTTP(DASH)is able to adjust video quality in real-time according to network communication.Considering its advantages,a stochastic optimization mechanism for DASH-based video service is proposed.Considering numerous important factors such as video bitrate,network energy consumption and quality of service experience,a video streaming service optimization problem is first established.Because the initial optimization problem is difficult to solve,the stochastic network optimization theory is further utilized to transform and decompose it into three easy-to-solve subproblems,including transmission control problem,bitrate selection problem and auxiliary computation problem.Based on the solutions of the three subproblems,a DASH-based video service optimization mechanism is established.Finally,the superiority of the proposed optimization mechanism is verified by simulation experiments using real PLR dataset.(4)Firstly,a newly emerged business model which achieves the cooperation between service provider and network operator is analyzed.Then a network service profit optimization framework is proposed in HSR network scenario which includes admission control,transmission control and resource purchasing control.Through the combination of millimeter-wave and microwave communication technologies,the communication bandwidth of HSR network is improved.Considering the video service data acceptance rate,differentiated video service quality,network service cost and data transmission rate,a profit optimization problem is built.Due to the complexity of the original problem,stochastic network optimization theory is utilized to decompose it into three subproblems,including admission control problem,transmission control problem and resource purchasing control problem.Based on the solutions of the three subproblems,a service profit maximization mechanism is proposed.Finally,through extensive simulation experiments,it is verified that the proposed mechanism can improve the profit of video service,ensure the differentiated service quality demands of mobile users,and effectively maintain the stability of HSR network.In summary,this thesis focuses on the key technologies of multimedia services in HSM networks,and systematically studies four research contents including data transmission,multi-base station selection optimization,video streaming service optimization and video service profit optimization.By organically organizing them into a complete research system,a whole research system is constructed and basic research results are output.In terms of research methods,this thesis utilizes a series of research methods such as related work research,mathematical modeling,algorithm design and experimental analysis.In data analysis,based on the real dataset,this thesis considers the actual performance index of HSM network and multimedia services,and verifies the superiority of the proposed method.The research results obtained in this thesis have certain significance for the research and development of multimedia service in HSM networks in the future.