Entry Date:
May 20, 2002

Multiparty Computation


Multiparty computation allows a group of players to perform a given task as correctly and as privately as if a trusted third party has performed the computation on players behalf. It was shown that any probabilistic (polynomial-time) function can be securely computed in the computational and in the information-theoretic settings. Ever since, many important papers have been produced and a lot of research is still going on in this fundamental area. In particular, the following issues are of critical importance: ability to compose secure protocols (e.g., sequential/parallel/concurrent reducibility), round and communication complexity, the network structure (e.g., existence of private channels, broadcast), adversary structure that can be tolerated (e.g., thereshold adversaries), types of adversary (e.g., active/passive/fail, adaptive/non-adaptive) and some others.