Font Size: a A A

Research On Streaming Media QoE Control Technology In Software Defined Network

Posted on:2019-06-26Degree:MasterType:Thesis
Country:ChinaCandidate:C SunFull Text:PDF
GTID:2428330542999652Subject:Electronics and Communications Engineering
Abstract/Summary:PDF Full Text Request
With the development of the mobile Internet,the number of live video platforms has grown dramatically,and people's demand for high-quality streaming media also increased.Streaming media transmission has a high sensitivity to network conditions,it is subject to link loss,delay,bandwidth,and other aspects,and then the transmission effect would be affected.The Software Defined Network(SDN),born at Stanford University,provides a new direction for the control of streaming media transmission.In the software-defined network,macro control and fine-grained control of the network are realized by separating to control plane and data plane.The controller and the switch use the southbound protocol OpenFlow to interact,and the controller control the network transmit behavior.This paper analyzes the generation and development of SDN,including the development of the Southbound protocol OpenFlow,the types of SDN transmission controllers,and the basic structure of the Ryu controllers used in this paper.This paper analyzes the development of streaming media transmission and quality of experience(QoE).With the increase of human-computer interaction requirements,quality of service(QoS)cannot reflect network transmission results.But the ultimate goal of streaming media transmission is to provide users with an acceptable QoE.The paper control streaming media transmission based on the SDN.It mainly aims at designing the best path based on the link loss in the media transmission.Furthermore,analysis the basic structure and basic principles of the SDN network,control the streaming media transmission based on QoE.The main work of the paper is divided into the following aspects:1.Set up the SDN network environment,implement the STP function to avoid broadcast storms,use the OpenFlow protocol to obtain the link loss of the network,design and implement a link loss-based streaming media transmission path selection system using the SDN-Dijkstra algorithm.The link loss is used as the weight to calculate the best path for transmission.The controller delivers the flow table to control the streaming media transmission.The transmission quality isevaluation by comparing the PSNR and SSIM between the source video and the received video,and use the MOS method to assess the QoE.2.Study the QoS-to-QoE mapping model,complete the mapping of QoS to QoE based on BP neural network,use the link loss,delay,jitter,and bandwidth as input,train multiple sets of data,further use deep beliefs network to establish a mapping model,and based on this model,designs a QoE-based streaming media transmission controller to obtain a better streaming media transmission effect.3.Using Docker to rapidly deploy SDN environment,using a hardware encoder board to encode a streaming media video source,using SDN network to transmit streaming media data,design and implement a windows application program,using a QoE-based controller to control Multi-channel video transmission and reception,finally get a better transmission effect.In summary,on the basis of the SDN and streaming media transmission technology,this paper studies the related technologies for streaming media transmission in SDN,designs the topology and control the network transmission in the Mininet platform.Designed a streaming media transmission display system.The streaming media transmission controller based on link loss and QoE proposed in the paper provides the basis for further research on streaming media transmission technology.
Keywords/Search Tags:SDN, streaming media, link loss, BP neural network, DBN, QoE
PDF Full Text Request
Related items