Font Size: a A A

Research On DAS-and SVC-oriented Transmission Optimization Technology In Named Data Networking

Posted on:2020-01-07Degree:MasterType:Thesis
Country:ChinaCandidate:W LiuFull Text:PDF
GTID:2428330578972251Subject:Computer application technology
Abstract/Summary:PDF Full Text Request
With the popularity of mobile terminals and the improvement of network infrastructure,people are no longer satisfied with simple session communication,but use the Internet to share multimedia content for most of the time.Nowadays,multimedia traffic has firmly occupied the dominant position of network traffic.The original intention of IP network design is to achieve end-to-end communication,which can not well meet the challenges of the current new trend of network traffic,and exposes a series of problems in mobility,content distribution,security and so on.In order to solve these problems completely,ICN(Information Center Network)is proposed.NDN(Named Data Networking)is a typical representative of ICN which has many features such as content-name-based routing,ubiquitous caching,natural support for multicast and receiver-driven transmission mode.It has great advantages in supporting large-scale repetitive video transmission.In recent years,more and more video websites adopt DASH(Dynamic Adaptive Streaming over HTTP)technology,which is a standard implementation of DAS(Dynamic Adaptive Streaming).It enables websites to distribute video content in an adaptive way and supports live broadcasting.Compared with other VOD(Video on Demand)and live broadcasting technologies,it has great advantages as well.In addition,Scalable Video Coding(SVC)is gradually being applied in the field of video because of its super adaptability to various heterogeneous devices and networks.The combination of NDN,DAS and SVC provides an effective solution for large-scale video distribution.Considering the features of DAS and SVC,we investigated the queue scheduling strategies and cache strategies of NDN router.The main contributions are as follows:Firstly,it was deeply studied that the layered characteristics of SVC video and the influence of the size of NDN request message and Data message on the forwarding process.Considering the natural convergence ability and ubiquitous caching features of NDN network,we proposed the NDN message queue scheduling strategies,BRF(Buffer and Reorder based Forwarding)and FBW(Forwarding based on Weight)that are suitable for transmitting SVC video by DAS technology.Video request message and low-level video data message are forwarded first,and this can effectively improve network throughput and content acquisition speed,thereby improving the quality of video playback and user experience.In addition,Pipeline's request mode and Smooth algorithm are added to the DAS client.They utilize the waiting time of client to send requests for high-level video blocks that are vacant in the buffer,so as to make full use of the network bandwidth and reduce the number of switches of the video playback quality,making the video playback smoother.Secondly,we studied the influences of cache time and location exert on cache hit rate and video playback quality,and proposed two SVC-oriented cache strategies,SRPC(SVC-oriented Reverse Probabilistic Caching)and SLPC(SVC-oriented Linear Probabilistic Caching).According to the hierarchical dependency of SVC coded video and the location of nodes in the transmission path,different caching probabilities are given to the data packets.The low-level SVC video blocks are cached near the client and the high-level video blocks are cached near the content source,which reduces the redundancy of network caching,increases the diversity of caching and improves the success rate of requests.Finally,we extended the ndnSIM module based on NS-3 environment,realized the improved DAS client,added two scheduling strategies and two cache strategies proposed by us to the NDN router,and verified their correctness and performance through simulation experiments.
Keywords/Search Tags:Named Data Networking, Dynamic Adaptive Streaming, Scalable Video Coding, Queue scheduling, Cache strategy
PDF Full Text Request
Related items