Modeling Research Of BitTorren File Pollution

BT (BitTorrent) is currently one of the most popular P2P file sharing systems. More and more users like the anonymity, openness and efficiency of BT, and many BT clients are widely used nowadays. In 2004, BT has occupied 53% of the entire P2P traffic, and in recent years, P2P traffic is 70% of the entire Internet traffic. So the BT security is becoming more and more important due to the prevalence, there are indeed many problems in BT network, file pollution is one needed more attention because of the urgency and harm, which is also one of the issues being debated.BT file pollution is including torrent pollution, index pollution and data pollution. Torrent pollution is simple, and we do not work on it. Index pollution and data pollution can both exist alone or combined with each other to form a more complex and subtle mode of pollution. The goal is to increase the probability of bad peer choice, to consume the bandwidth and connections of other nodes, and finally to reduce the download speed and let the network resource become unavailable. The index pollution and data pollution technology are relatively complex, and are also lack of systematic research methods and theories, we focus on the index of pollution and data pollution to analysis model for the BT network file pollution, to provide a scientific reference of BT network.In this paper, we introduce the BitTorrent protocol, the background knowledge and analysis the files polluting technologies. And then introduce several common models and pollution resistance models, analyze the focus of these models, the advantages and how to use, and the pointed out that this effective pollution model must take into consideration. Then we propose a combination model of index pollution and data pollution, which combines the blacklist mechanism, the node selection strategy and other key factors of the effect of pollution, and introduced the concept of parallelism within the block, model the effect of pollution and pollution costs. Finally, we design and implement a system for high-performance file pollution, contaminated by a large number of experimental results, we analysis and verify the correctness of the hybrid pollution model, and calculate the indicator values of effect of pollution and pollution costs.