P2P Traffic Identification Based On Trust Sampling

Although P2 P technology have promote the vigorous development of the Internet, but P2 P traffic had caused a great impact on the traditional Internet architecture such as the backbone of the central router. According to statistics: In the past few years, P2 P applications on the network bandwidth usage than in the backbone network of approximately 40%-70%, in the metro or LAN may even exceed 80%, consumes a lot of network bandwidth. P2 P traffic identification of significance for the Internet service provider(ISP) to better manage the network, reduce network operation and maintenance costs, improve network quality of service provided data analysis and monitoring direction. The nature of P2 P traffic identification is to identify all of the nodes under the ISP network router where the proportion of P2 P traffic in a certain time period, the trust-based sampling P2 P traffic identification system, This paper focuses on the trust-based sampling of P2 P traffic identification, the main work as follows:(1) P2 P traffic identification based on first-state trust sampling frame. according to the proposed modular architecture and coding trust sampling strategy to achieve a simple first-order trust sample identification system, and the system under test and experimental data in the environment are analyzed and summarized.(2)The research on identification based on big data platform. a large data processing platform parallel algorithms which the muti-pattern match algorithms run on, and a brief description of the algorithm and data packet messages packaging.(3)Second-state trust sampling P2 P traffic identification framework. Stage sampling algorithm proposed reservoirs and analyze the feasibility of the algorithm and prove the correctness of the algorithm. Finally, the simulation of the policy, and the simulation results are analyzed and summarized.