Research On Effective Transmission Technology Of Multimedia Service Based On Network Coding

The development trend of Internet services shows that the multimedia service will definitely be the dominating and primary one in the future Internet. So how to optimize network architecture and algorithm designment and therefore improve the performance of service guarantee for multimedia service is one of the most important problems faced by the current Internet. Focused on this item, this paper researches how to make network coding theory more suitable for efficient transmission of multimedia and then designs and optimizes both network and node for efficient transmission of multimedia. On the one hand for network optimization, this paper focuses on the decrease of multimedia transmission delay based on network coding and then improves the bandwidth usage and transmission efficiency; On the other hand for node design, this paper studies on efficient switching fabric and the related scheduling algorithm for better transmission of multimedia services. At last, this paper provides an optimization method for the congestion control of network coding in efficient transmission by gradient projection. In detail, the main research achievements of this paper are as follows:Focused on the decoding delay problem for network coding, this paper provides a low-delay network coding constructing (LDNCC) strategy based on the multimedia services. This strategy chooses one of the multiple disjoint paths between source node and destination node as the expedient one for the transmission of un-coded packets, and therefore decreases the decoding delay. For low decoding delay of destination node, the LDNCC stategy provides the optimal expedient paths selection algorithm to find more expedient paths so as to make sure of effective multimedia transmission. The simulation shows that LDNCC stragety could decrease decoding delay efficiently while keeping the throughput of network coding, and therefore is an ideal solution for delay sensitive services such as multimedia compatibly.For the multi-source multicast scene, this paper puts forward a multi-source layered multicast throughput optimization(MSLMTO) strategy based on network coding. The strategy provides the genetic algorithm (GA) and improving genetic algorithm (GA+) for the optimal network coding of multi-source layered multicast. Further more, the strategy could converge to the optimal solution fast by designing reasonable chromosome structure and generation method of initialization swarms. The simulation shows that compared with traditional layered multicast strategies, MSLMTO strategy could improve the throughput of multi-source layered multicast and then take advantage of network resources for the transmission of multimedia services.Focused on the delay rising problem leading by the nodes cumulation in traditional network coding when increasing node throughput, this paper proposes a layered cascaded fusion coding scheduling(LCFCS) scheme based on multimedia services. In the condition of users' Quality of Experience(QoE), this strategy make the fusion of information coding and network coding, in order to both compress transmission packets of multimedia services by information coding and reduce the nodes switching presure by network coding. The fusion of information coding and network coding can improve the efficiency of switching nodes. To make the maximum throughput of switching nodes,this strategy takes the dependence of random distribution of access time into consideration and makes the minimum of both cost and link flows as the objective. Then this paper provides the LCFCS mathematical model and maximal coloring(MC) algorithm to search the maximal stable sets for the LCFCS model solution. The simulation shows that the model and related algorithms perform well for efficient transmission and delay guarantee.Current switching mechanisms fall short of multicast in QoS and throughput guarantee for multimedia services. This paper makes researches on implementation of linear network coding into combined input and cross-point queuing switches based on the analysis of linear network coding theory, and then proposes a high throughput guaranteed multicast switch mechanism for multicast services which is named Most Urgent First Coding (MUFC). Compared with the current network coding switching mechanism based on CICQ, MUFC could switch packets from arbitrary input and output ports independently with high scalability. The theoretical analysis and simulation results show that MUFC would obtain high throughput and favorable delay performance without speedup which could improve the transmission ability for multimedia services well.Focused on the congestion control optimization problem for network coding, provides a congestion control method for network coding based on gradient projection optimization which is named NCGPO. NCGPO could optimize the congestion while guarantee the reliability by dynamicly adjust traffic allocation besides lower the encoding rate. Based on the analysis of the essence of data transmission by network coding, this paper models the congestion control progress of NCGPO, and takes the gradient projection into consideration to optimize the model. The stability of the algorithm is also proved. The simulation shows that the method could optimize the congestion and network resource as well as construct low-delay network coding. The NCGPO could balance the load of both the link and the node, which could transmit multimedia service efficiently. So to sum up, the main innovations of this paper are as follows:Presenting a low-delay network coding constructing(LDNC) strategy;Puts forward a multi-source layered multicast throughput optimization(MSLMTO) strategy based on network coding;Proposing a layered cascaded fusion coding scheduling(LCFCS) model and a maximal coloring(MC) algorithm to search the maximal stable sets for the model solution.Proposing a high throughput guaranteed multicast switch mechanism for multicast services Most Urgent First Coding(MUFC) based on the combined input and cross-point queuing switches and linear network coding theory.Provides a congestion control method for network coding based on the gradient projection optimization which is named( NCGPO).