BFR-MPC: A Blockchain-Based Fair and Robust Multi-Party Computation Scheme

Abstract

In a general secure multi-party computation (MPC) protocol, two or more participants who do not trust each other, use their respective secret inputs to calculate a joint function in a distributed environment without a centralized organization. They can get correct outputs on the premise of ensuring privacy and independence of input. In this paper, to solve the problem of fairness and robustness in MPC, a blockchain-based multi-party computation scheme (BFR-MPC) was proposed. The blockchain maintains an open reputation system for parties as a public ledger where a more reputable party has a greater chance to be selected. The block height is used as a trusted timestamp. In each round, parties must send the correct information before the deadline. In our scheme, all parties are considered to be foresighted, and an incentive mechanism encourages parties to cooperate rather than deviate from the protocol. Because of non-cooperative parties will be immediately expelled from the protocol and will be penalized financially, the proposed scheme is robust. The penalty will be used to reward honest parties. We also proved the fairness of our scheme through Game Theory. The comparison results of the proposed scheme with other schemes show that it is a more practical scheme for MPC with high fairness and robustness.

Keywords

• Multi-party computation
• blockchain
• smart contract
• fairness
• robustness
• secret sharing