The front line battle against P2P

Abstract

One of the prevalent topics under discussion in educational environments today is the use of P2P (peer-to-peer) software. Aside from being major bandwidth-eaters, the use of such software to distribute and download copyrighted material has significant legal and financial implications for the campus community. Having a clear and concise policy restricting such use is one matter; enforcing said policy has proved to be an entirely different and complex task. Although it is elementary to prevent P2P traffic by blocking well-known ports used by such software, the vast majorities majoritymof P2P software, such as BitTorrent, download and upload pieces of files from different sources on different ports[?]. P2P protocols and client programs are working around 'issues' that prevent them from functioning and are becoming more and more 'intelligent'. The detection of P2P traffic based on well-known ports and 'port guessing' do not match the intelligence in newer generation P2P protocols and software. False positives generated with the use of well-known ports for detection are another serious concern. A good approach would be a solution that detects P2P and other undesirable traffic based on packet payload, as opposed to source and destination ports. Commercial products available to address these issues are expensive and not as flexible as the solution we discuss in this paper. Our proposed model is a combination of several open-source solutions to which changes have been made to suit our environment and requirements. We also discuss other possible model solutions and discuss the pros and cons of such solutions.