An efficient approach for secure multi-party computation without authenticated channel

Abstract Secure multi-party sum is one of the most important secure multi-party computation protocols. It has been widely applied to solve many privacy-preservation problems such as privacy-preserving data mining, secure auction, secure electronic voting, and privacy-preserving statistical data analysis. To guarantee the correctness of the final output and to enhance the security level, the existing secure multi-party sum protocols have to use authenticated channels, even secure channels for participants to communicate, however such a usage requirement significantly reduces their performance. Furthermore, these secure multi-party sum protocols are impossible to run on public networks. In this paper, we propose a new secure multi-party sum protocol that can ensure the correctness of the output result as well as securely protecting the parties’ privacy against attacks without requiring any authenticated/secure channel. The proposed protocol is based on a multi-party sum function employing a variant of ElGamal encryption and a Schnorr signature-derived authentication method, in which both these cryptographic tools use the same private and public keys. Additionally, our comparative evaluation shows that the proposed protocol is efficient and practical.