[论文解读] Correctness and Fairness of Tendermint-core Blockchains

Yackolley Amoussou-Guenou, Antonella Del Pozzo+2 more|arXiv (Cornell University)|May 22, 2018

Blockchain Technology Applications and Security参考文献 31被引用 21

一句话总结

本文形式化验证了Tendermint-core区块链的正确性与公平性，证明在最终同步环境下，若拜占庭验证者少于1/3，经修改的协议版本可实现一次性及重复性共识。研究建立了一个公平奖励机制存在的充要条件：仅当系统为最终同步时成立；并表明原始Tendermint奖励机制不公平，但通过优化超时机制和消息内容可使其变为最终公平。

ABSTRACT

Tendermint-core blockchains (e.g. Cosmos) are considered today one of the most viable alternatives for the highly energy consuming proof-of-work blockchains such as Bitcoin and Ethereum. Their particularity is that they aim at offering strong consistency (no forks) in an open system combining two ingredients (i) a set of validators that generate blocks via a variant of Practical Byzantine Fault Tolerant (PBFT) consensus protocol and (ii) a selection strategy that dynamically selects nodes to be validators for the next block via a proof-of-stake mechanism. However,the exact assumptions on the system model under which Tendermint underlying algorithms are correct and the exact properties Tendermint verifies have never been formally analyzed. The contribution of this paper is two-fold. First, while formalizing Tendermint algorithms we precisely characterize the system model and the exact problem solved by Tendermint. We prove that in eventual synchronous systems a modified version of Tendermint solves (i) under additional assumptions, a variant of one-shot consensus for the validation of one single block and (ii) a variant of the repeated consensus problem for multiple blocks. These results hold even if the set of validators is hit by Byzantine failures, provided that for each one-shot consensus instance less than one third of the validators is Byzantine. Our second contribution relates to the fairness of the rewarding mechanism. It is common knowledge that in permisionless blockchain systems the main threat is the tragedy of commons that may yield the system to collapse if the rewarding mechanism is not adequate. Ad minimum the rewarding mechanism must be fair, i.e.distributing the rewards in proportion to the merit of participants. We prove, for the first time in blockchain systems, that in repeated-consensus based blockchains there exists an (eventual) fair rewarding mechanism if and only if the system is (eventual) synchronous. We also show that the original Tendermint rewarding is not fair, however, a modification of the original protocol makes it eventually fair.

研究动机与目标

• 在现实系统假设下，对Tendermint-core共识协议的形式化分析。
• 识别并修复原始Tendermint协议中影响共识安全性和活性的细微错误。
• 基于验证者贡献度，定义并证明重复共识区块链中奖励机制的公平性。
• 确立区块链系统中最终公平奖励机制存在的必要与充分条件。
• 弥合实际区块链实现与形式化分布式系统理论之间的鸿沟。

提出的方法

• 使用带有最终同步假设的分布式系统框架，对Tendermint-core协议进行形式化建模。
• 引入优化的超时机制，以修正原始协议中的活性问题。
• 定义一种基于贡献度的奖励机制，其中验证者根据其正确参与和消息传递的比例获得奖励。
• 在少于1/3的验证者为拜占庭的前提下，证明共识终止与有效性。
• 采用修改后的共识模型，使预提交消息包含可区分信息，以识别正确与拜占庭进程。
• 通过双向蕴含证明，确立最终公平奖励机制存在的充要条件：系统为最终同步时成立。

实验结果

研究问题

• RQ1 在何种系统假设下，原始Tendermint协议能实现共识正确性？
• RQ2 原始Tendermint协议能否通过形式化方法修正，以确保在存在拜占庭验证者时仍具备活性与安全性？
• RQ3 重复共识区块链中奖励机制被视为公平的必要与充分条件是什么？
• RQ4 在异步系统中，Tendermint-core区块链的奖励公平性是否可能实现？
• RQ5 原始Tendermint奖励机制能否被修改以实现最终公平性？

主要发现

• 在最终同步系统中，若少于1/3的验证者为拜占庭，经修改的Tendermint协议可实现一次性及重复共识。
• 原始Tendermint协议在标准假设下不正确，因其存在活性错误，但可通过优化超时机制修复。
• 在重复共识区块链中，公平奖励机制存在的充要条件是系统为最终同步或同步系统。
• 原始Tendermint奖励机制即使在最终同步系统中也不公平，原因是在同步过程中消息可见性不完整。
• 通过接收超过2n/3条消息并优化超时设置，可实现改进的奖励机制，使其在最终同步系统中达到最终公平。
• 在异步系统中，即使所有验证者均正确，也不存在（最终）公平奖励机制，证明了此类环境中公平性的根本局限。