Research On Scheduling Mechanism And Performance Evaluation For SDN-oriented Video Streams

Due to the popularization of the Internet application,the volume of online video services on the Internet has grown exponentially within recent years.It is normal for Internet users to have access to online-video services such as online TV,online gaming,on-demand video,and video conferencing.Video services bring higher requirements for traditional network nodes including higher video processing capability,greater throughput,and lower latency networks.The current network architecture is hard to meet the future needs of video services.Therefore,there is an urgent need to improve the existing network video processing capacity and to guarantee the service quality of different video services with a more flexible and effective network architecture.This paper proposes a multipriority Fork-Join(MPFJ)scheduling model under the SDN network architecture,which can guarantee the development of multimedia video services with effective utilization and allocation of network resources.The main contributions of this article are as follows.(1)According to the characteristics and different requirements of the current multimedia video service,a multiuser request classification algorithm is proposed,which takes the QoS requirements of different video services into consideration.Combining the priority-based algorithm with the Fork-Join parallel scheduling and the non-preemptive scheduling,a multipriority Fork-Join(MPFJ)scheduling algorithm is proposed.The MPFJ algorithm has the ability to meet the new requirements of video services with a SDN network environment,improves the service capabilities of the network,and guarantees the quality of service(QoS)of different video services.(2)Theoretical analyses are performed using stochastic network calculus to investigate the behavior of Fork-Join scheduling model and MPFJ(multipriority Fork-Join)scheduling model.The MPFJ algorithm based on the proposed model can be applied to improve the network quality of services.The network performance boundaries are obtained using network calculus to study the limits of backlog and delay.These theoretical investigations are constructive in improving the QoS expectations of video streaming with the proposed algorithms.(3)Numerical simulations of the proposed MPFJ algorithm using OMNeT++ simulation test bed are conducted while a SDN network is established accordingly to verify proposed model and the proposed algorithm for video streaming.Detailed comparisons and analyses are conducted based on the data from the SDN network.Starting with the local network node,this paper studies the End-to-end performance of a network system including multiple switches.Considering the QoS requirement of different video service users,the new scheduling model can improve the service ability of network system in video streaming by combining the multiple priority scheduling and forkjoin scheduling mechanism.Finally,by applying the stochastic network calculus theory,the performance are analyzed by the video stream of the network system,which provides the theoretical foundation for the video data flow through multiple switches.