EquiBFT: A Framework for Achieving Fairness in BFT Consensus

Abstract

Byzantine Fault-Tolerant (BFT) consensus protocols are increasingly utilized in blockchain environments. In such protocols, the leader node holds the authority to dictate the transaction order, potentially impacting the fairness of decentralized finance (DeFi) applications. For instance, attackers can exploit this to manipulate transaction order and conduct front-running attacks. The concept of order-fairness, which recently emerged, has become a critical property for preventing a single node from unilaterally determining transaction order. Protocols designed to uphold order-fairness often rely on the sequence in which transactions appear across the network, a factor that can be influenced by the network’s topology. However, this approach has inherent limitations, such as challenges in avoiding Condorcet cycles (Kelkar et al., Crypto 2020).To address these challenges, we propose a novel definition of fairness that requires concealing transaction content before ordering. Additionally, we extend the definitions of liveness and safety of consensus protocols to cover the transaction decryption process, guaranteeing the successful decryption of transactions. Based on the existing BFT protocol and utilizing threshold encryption algorithms, we designed a framework called EquiBFT which can incorporate fairness to BFT protocols. We have proven that the EquiBFT satisfies fairness while ensuring the liveness and safety. We implemented this framework based on HotStuff (Yin et al., PODC 2019) and validated its feasibility in a real-world network environment.