This pages presents research results relating to secure multicast that have been discovered in the Cryptography and Information Security Group of MIT's Lab for Computer Science.

Multicast communication is an attractive method for delivery of data to multiple recipients, minimizing consumption of both sender and network resources. Multicast is supported, for example, on the internet, or via satellite communication. Some applications that benefit from multicast include real-time information update, multi-party conferencing, and pay TV. Securing multicast communication poses several important challenges. In the paper


• Efficient Communication-Storage Tradeoffs for Multicast Encryption by Ran Canetti, Tal Malkin, and Kobbi Nissim. Eurocrypt 99.
  
  we focus on providing access control, namely ensuring that only legitimate members of the multicast group have access to the group communication. This is done by maintaining a secret session key known to all legitimate members, and encrypting all group communication using this key. We consider a dynamic group, where users join or leave the group in an arbitrary fashion, and a center who is in charge of performing the re-keying associated with group updates. We require strong security, where the session key is secure against any coalition of nonmembers.

  There is a variety of different scenarios using multicast, presenting a range of efficiency requirements with respect to several parameters. We give an upper bound on the tradeoff between storage and communication parameters. In particular, we suggest an improvement of the schemes by Wallner et al. and Wong et al. with sub-linear center storage, without a significant loss in other parameters. Correctly selecting the parameters of our scheme we can efficiently accommodate a wide range of scenarios. This is demonstrated by applying the protocol to some known benchmark scenarios.

  We also show lower bounds on the tradeoff between communication and user storage, and show that our scheme is almost optimal with respect to these lower bounds.