Synchronous Consensus in Partial Synchrony

TL;DR

The paper addresses consensus in the presence of partial synchrony by introducing a deterministic Byzantine consensus algorithm that operates synchronously in a partial-synchrony environment. It achieves $f< n/2$ resilience with only $2$ messaging rounds and leverages $3$-hop epidemic delivery to circumvent asynchronous/faulty links, enabling bounded termination and simultaneous validity with a leaderless design. The authors provide formal proofs of tolerance to link asynchrony, a Trim Rule for consistent decision sets, and a two-phase protocol using securely signed Initial Value and Agreement Proposal messages. Practically, this advances non-blocking distributed transactions and fault-tolerant replication in decentralized systems by offering a scalable, leaderless consensus mechanism with provable liveness guarantees under partial synchrony.

Abstract

We demonstrate a deterministic Byzantine consensus algorithm with synchronous operation in partial synchrony. It is naturally leaderless, tolerates any number of $ f<n/2 $ Byzantine processes with 2 rounds of exchange of originator-only signed messages, and terminates within a bounded interval of time. The algorithm is resilient to transient faults and asynchrony in a fraction of links with known size per number of faulty processes. It circumvents asynchronous and faulty links with 3-hop epidemic dissemination. Key finding: the resilience to asynchrony of links and the enabled by it leaderless consensus in partial synchrony ensure algorithm operation with simultaneous validity, safety, and bounded liveness.

Figures (5)

• Figure 1: Consensus with 2 Faulty of 5 Processes
• Figure 2: Consensus with 1 Faulty of 5 Processes
• Figure 3: Consensus with 0 Faulty of 5 Processes
• Figure 4: Tolerance to Async/Faults (Computation Results)
• Figure 5: Tolerance to Async/Faults (Computed + Analytical)