Measuring and characterizing peer-to-peer (P2P) file-sharing systems will benefit the optimization and management of P2P systems. Though there are a lot of measurement studies on BitTorrent almost in every important aspect, few of them focus on the measurement issues and the corresponding solutions, which can strongly influence the accuracy of measurement results. This paper analyzes the key difficulties of measuring BitTorrent and presents a measurement system with combination of active and passive ways, which can han-dle with the problems well and balance the efficiency and integrity. Then compared to other work, a more complete and representative measurement was performed for nearly two months and several characteristics are concerned: 1) there are diverse content sharing in BitTorrent system, but multimedia files that are larger than 100 MB are the most. 2) Distributed Hash Tables has indeed enhanced the ability of peer discovery though there are some pitfalls to be addressed. 3) Pieces are distributed uniformly after the early stage and there are few rare pieces. Furthermore, peer arrival rate shows a periodical pattern, which was not well mod-eled before. Then an improved model is proposed and the experiment results indicate that new model is fitted in with actual measurement results with high accuracy
SummaryIllegal sharing of copyrighted materials in peer‐to‐peer (P2P) networks has become a serious threat to the content industries. Despite many copyright‐protected P2P systems that have been designed to prevent piracy, the user authentication–based approaches face the challenge of collusion attacks and the encryption‐based approaches may be vulnerable to poisoning attacks. Observing these, in this paper, we present a novel copyright‐protected content delivery scheme (CPP) for P2P file‐sharing networks based on piece‐level stochastic encryption. In CPP, peers must pay for decryption key sequences before accessing to the plaintext content. For each peer, the sequence of the decryption keys is different from others, so the unauthorized peers could not access the content even if some colluders share their keys. Modified piece with hash enable peers to check the correctness of each piece after downloading, preventing the propagation of poisoned pieces. There is neither massive state maintenance nor frequent user authentication during the download process. Peers purchase after downloading, protecting their benefits when quitting the system before paying. Analysis and experiments show that CPP is suitable for delivering large files, and it is resilience to collusion attacks and poisoning attacks.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.