Research On Coding And Transmission To Guarantee Quality Of Service In P2P Streaming Media Content Delivery Networks

In recent years, thanks to rapid development of Internet and streaming media technology, contents of Internet transmission have changed from simple texts and graphics to streaming media data combining text, audio and video. This change demonstrated that more and more users are expecting more intuitive and diversified information, and marked the arrival of streaming media age of Internet. Network streaming media technology is becoming one of the most popular applications on Internet and mobile network. Because it is entertaining and interactive, streaming media is becoming major incentive for the development of broadband network and mobile network. However, because it is faced with the demand of large-scale streaming media distribution service, how to utilize limited resources, satisfy the increasing need of the users, and reduce the cost while guarantee QoS (Quality of Service), has become challenging tasks in researches of streaming media distribution technology.Content distributing network of streaming media is aimed at improving speed of accessing streaming media content, establishing and covering streaming media application network. This network is specialized in efficient transmission of streaming media content, which can reduce the load of source server, mitigate network congestion, and improve the usability of the system. However, this network need high expenses on distribution and maintenance and is limited by handling capacity of edge server. In order to guarantee QoS of streaming media, we utilize the autonomy and extendibility of P2P network to reduce the cost of distribution, and put forward framework of mixed-content distribution system based on P2P streaming media.In order to improve the performance of mixed-content distribution system based on P2P streaming media, we reduced the consumption of network resources during transmission of streaming media data. Based on characteristic of application of PSP streaming media, we studied scalable coding scheme, and put forward multiple-description layered coding technology based on pixel space decomposition, which can guarantee heterogeneity and reliability of P2P network. Meanwhile, we introduced video sub-layer merging algorithm and pixel estimation algorithm, which can improve QoS of user node video. By analyzing the relations between quantized values and estimating error of pixels, we established multi-description layered coding and decoding mathematical model. The scheme has some advantages regarding to heterogeneity of P2P network, and can ensure reliability of network transmission.Content distribution network of P2P streaming media is multicast and unstable. In order to improve reliability and safety of streaming media, we have studied network coding technology, and proposed an improved stochastic linear network coding scheme based on P2P stream media by utilizing the characteristics of multiple-description layered coding and network coding. And then we did simulation experiment on the PSP simulation platform GPS, compared node upload and download of network coded PPVOD system and ordinary BitTorrent system, the result shows network coded PPVOD has advantages in node upload and download.According to "best-effort" service model of network, we suggested RSVP resource reservation mechanism based on application layer, which can guarantee QoS of streaming media application. We have designed corresponding message and message mechanism, which can improve effective utilization rate of resources on P2P content distribution system of streaming media. Through simulation experiment and comparison of performance of system before and after using resource reservation mechanism based on application layer, we got the conclusion that resource reservation mechanism can improve the overall performance of the system.According to the nature of network, such as heterogeneity of terminals and limited bandwidth resource, we suggested LMDBA, a dynamic bandwidth distribution strategy based on multiple-description layered coding. This strategy has achieved dynamic distribution of bandwidth through using multiple-description layered coding, and applied QoS upgrading and downgrading strategy. This strategy can mitigate systemic blockage, improve utilization rate of bandwidth resources while ensure QoS. We also have done equity analysis of QoS upgrading and downgrading strategy. The result of simulation experiment shows that the new strategy can significantly improve utilization rate of bandwidth resource and mitigate systemic blockage, compared with traditional fixed-bandwidth distribution strategy, while QoS is ensured.