SoK: Consensus for Fair Message Ordering

Abstract

Distributed ledger systems, such as blockchains, rely on consensus protocols that commit ordered messages for processing. In practice, message ordering within these systems is often reward-driven. This raises concerns about fairness, particularly in decentralized finance applications, where nodes can exploit transaction orders to maximize rewards referred to as Maximal Extractable Value. This paper provides a systematic understanding of consensus protocols that order messages with different approaches, especially focusing on the ones that promote order fairness, using methods including First-In-First-Out (FIFO), random, and blind ordering. We review the challenges and trade-offs of deriving fair message ordering in a Byzantine fault-tolerant setting, and summarize the requirements for making a fair message ordering consensus protocol. We introduce a design guideline, with which we propose a latency optimization to the state-of-the-art FIFO ordering protocol of Themis. This work provides a systematic way for assessing and enhancing message order fairness in blockchain systems.