Randomized Traitor Tracing Algorithms for Broadcast Security

We introduce and analyze a randomized traitor tracing algorithm for broadcast security. This algorithm is very efficient compared to brute force methods (the mean time it takes is O(√T ) where T is the mean time taken by the brute force algorithm) and can catch traitors, i.e., dishonest users who collude by giving subsets of their key(s) to others. The system model we use can be applied to various broadcast scenarios, e.g., over the internet or in a PayTV system. We also prove a structure theorem for the “square” codes we design for traitor tracing.