With many torrent users worldwide and MPAA/RIAA being very active, at the first glance I thought the paper was funded by them.
The paper being "BITSTALKER: ACCURATELY AND EFFICIENTLY MONITORING BITTORRENT TRAFFIC" from University of Colorado. Boulder CO.
The researchers claim that their tool is more accurate and will eliminate false positives. But people involved with torrent sites and operators seem to be less impressed by the technology according to Torrent Freak.
Like torrent freak we too wonder why ISP's, major US ISP's at that, investing in this arena, which has been the field of RIAA/MPAA. ISPs usually remained neutral to the witch hunting coducted by them.
Following is an excerpt from the paper and for more descriptive account of BitStalker and positives and negatives of it at Torrent Freak, and slashdot (link after the jump)
This paper presents BitStalker, a low-cost approach to monitoring large BitTorrent file sharing swarms. BitStalker collects concrete evidence of peers’ participation in file sharing in a way that is robust to tracker pollution, highly accurate, and efficient. In contrast, the past and present investigative monitoring strategy consists of tracker server queries and ICMP ping probes. While this method is simple, it is also prone to a variety of significant errors, especially false positive identification, since this monitoring technique does not verify participation in the file sharing. We present an alternative monitoring strategy based on actively probing the list of suspected peers to obtain more conclusive evidence of participation in the file sharing. There are several aspects of our approach that warrant additional attention. In particular, a specific definition of what constitutes “evidence” in the context of file sharing across various legal systems should be explored. Also, the general legal issues that this type of monitoring exposes should also be investigated further.Plycypher is a a broadband ISP consortium (COMCAST, COX and Time Warner Cable)
Acknowledgments. The authors thank the anonymous reviewers for their valuable comments and suggestions. We also thank Claire Dunne and the University of Colorado’s institutional review board for ensuring that this research was conducted with the highest of ethical standards. This research was funded in part through gifts from PolyCipher.