| The analyses based on large-scale simulation is a technical means for the evaluation of CoolStreaming and other network protocol and algorithm. CoolStreaming is a P2P live-streaming protocol based on data-driven mechanism. Data-driven mechanism means that it is the availability of data that guides the flow directions, not a specific overlay structure that restricts the flow directions. The Peer-to-Peer (P2P) networktransfer can observably reduce the payload of the source peer and improve the utilization of the network resources by sharing the bandwidth and resources between peers.The thesis mainly discusses the implementation and analyses of Coolstreaming in parallel simulation. The network simulation platform NS2 and it's parallel version PDNS have been extended. The platform supplies with the configuration interface for interactions in the protocol among peers, and also the method for data output and statistic. The platform simulates the actions of peers and source peer and implements the transmission of control message and live-streaming data. After the extension, it's feasible to bring out large-scale parallel simulation on the PDNS platform.Many large-scale simulation analyses are performed on the extended platform. It tests the function and performance of the program. Then according to the key parameters such as continuity, delay-before-play, rate of process and network traffic load, the thesis assesses the application of the protocol and network performance,finds the rational parameters and the key factors that effects the performance, at last shows the comparison of the improvement of partial algorithm.It shoud be mentioned that the results presented in this thesis can be effectively applied to further evaluation of CoolStreaming and new P2P protocols and algorithm in the future. |
PDF Full Text Request