The objective of Peer-to-Peer Content Delivery Networks is to deliver copyrighted contents to paid clients in an efficient and secure manner. To protect such contents from being distributed to unauthorized peers, Lou and Hwang proposed a proactive content poisoning scheme to restrain an illegal download conducted by unauthorized peers, and a scheme to identify colluders who illegally leak the contents to such unauthorized peers. In this paper, we propose three schemes which extend the Lou and Hwang's colluder detection scheme in two directions. The first direction is to introduce an intensive probing to check suspected peers, and the second direction is to adopt a reputation system to select reliable (non-colluder) peers as a decoy. The performance of the resulting scheme is evaluated by simulation. The result of simulations indicates that the proposed schemes detect all colluders about 30% earlier on average than the original scheme while keeping the accuracy of the colluder detection at medium collusion rate.------------------------------ This is a preprint of an article intended for publication Journal of Information Processing(JIP). This preprint should not be cited. This article should be cited as: Journal of Information Processing Vol.20(2012) No.2 (online) DOI　http://dx.doi.org/10.2197/ipsjjip.20.378------------------------------

The objective of Peer-to-Peer Content Delivery Networks is to deliver copyrighted contents to paid clients in an efficient and secure manner. To protect such contents from being distributed to unauthorized peers, Lou and Hwang proposed a proactive content poisoning scheme to restrain an illegal download conducted by unauthorized peers, and a scheme to identify colluders who illegally leak the contents to such unauthorized peers. In this paper, we propose three schemes which extend the Lou and Hwang's colluder detection scheme in two directions. The first direction is to introduce an intensive probing to check suspected peers, and the second direction is to adopt a reputation system to select reliable (non-colluder) peers as a decoy. The performance of the resulting scheme is evaluated by simulation. The result of simulations indicates that the proposed schemes detect all colluders about 30% earlier on average than the original scheme while keeping the accuracy of the colluder detection at medium collusion rate.------------------------------ This is a preprint of an article intended for publication Journal of Information Processing(JIP). This preprint should not be cited. This article should be cited as: Journal of Information Processing Vol.20(2012) No.2 (online) DOI　http://dx.doi.org/10.2197/ipsjjip.20.378------------------------------