GUC-Secure Join Operator in Distributed Relational Database

Privacy-preserving SQL computation in distributed relational database is one of important applications of secure multiparty computation. In contrast with comparatively more works on privacy-preserving data-query in database, only few works deal with provably-secure privacy-preserving data manipulations, among which the join operator is the most powerful in generating new data (relation). We propose a very general cryptographic protocol framework for secure 2-party join computation based on anonymous IBE (identity-based encryption) scheme and its user private-keys blind generation techniques. This construction is provably GUC (generalized universally composable) secure in standard model with acceptable efficiency. In addition, an efficient instantiation based-on the anonymous Boyen-Waters IBE scheme is presented which user private-key’s blind generation protocol may be of independent values.