Leaker Identification in Multicast Communication

Abstract

Multicast is a communication mode in which data is exchanged among multiple end systems. An important concern in multicast communications is the protection of copyrights. While preventing copyright violations seems very difficult, tracing copyright violators (leaker identification) is more feasible and can be used as a deterrence alternative. In order to identify leakers in a multicast environment, every receiver should obtain a uniquely marked copy of the data. However, delivering unique copies from the sender to the receivers is inefficient. This paper investigates multicast leaker identification and introduces an efficient solution that is based on binary search tree. We introduce the notion of suspicious set that includes suspected end systems. When a leak is detected, all receivers are inserted in the suspicious set. Using a binary search, the suspicious set is refined successively until it includes only the leakers. An analytical study is conducted to evaluate the proposed solution.