Design And Simulation Of P2P Live Streaming Protocol Based On SVC Streams

Owing to the high scalability and features of resource sharing, Peer-to-Peer networksachieved great success in the field of video live streaming, widely favored by academiaand industry. But now the uncertainty of network bandwidth, the diversity of the terminalcapability and so on, all required the P2P streaming system provide a suitable quality ofservice to each peer based on their actual need. In order to support such scalable mediaservice, Scalable Video Coding was proposed. Therefore, by the real-time transmissionof SVC streams, P2P systems could support the scalable live streaming service,orchestrating the reasonable delivery, adapting to the nerwork bandwidth and terminalequipment, which has a very broad application prospects now.This thesis first studys the key technology of P2P live streaming algorithm design,and related research works specialized for the SVC streams transmission, analyzes theadvantages for the realization and deployment of a data-driven P2P live streamingsolution in the heterogeneous environment. Through further study of the SVC technology,the thesis extends the existing the reference software and framework to NS2environment,using the real bit stream to simulate and evaluate the H.264/SVC transmission.According to data structure of H.264/SVC coded video, the thesis analyzes thecharacteristics of the pull-based data scheduling in a heterogeneous environment andimprove the P2P data scheduling algorithm, and finally puts forward the wholeframework and simulation realization of the P2P live streaming protocol with scalablequality of service. By the comparative analysis and evaluation of the scale scalability,transmission effiency, continuity and quality, the simulation results demonstrate highadaptability of the proposed algorithm for the heterogeneous environment, in line withthe theoretical analysis.