Pierre Civit, Muhammad Ayaz Dzulfikar, Seth Gilbert, Vincent Gramoli, Rachid Guerraoui, Jovan Komatovic, Manuel Vidigueira
{"title":"拜占庭共识是 $$\\Theta (n^2)$$:即使在部分同步的情况下,Dolev-Reischuk 定界也是紧密的!","authors":"Pierre Civit, Muhammad Ayaz Dzulfikar, Seth Gilbert, Vincent Gramoli, Rachid Guerraoui, Jovan Komatovic, Manuel Vidigueira","doi":"10.1007/s00446-023-00458-w","DOIUrl":null,"url":null,"abstract":"<p>The Dolev-Reischuk bound says that any deterministic Byzantine consensus protocol has (at least) quadratic (in the number of processes) communication complexity in the worst case: given a system with <i>n</i> processes and at most <span>\\(f < n / 3\\)</span> failures, any solution to Byzantine consensus exchanges <span>\\(\\Omega \\big (n^2\\big )\\)</span> words, where a word contains a constant number of values and signatures. While it has been shown that the bound is tight in synchronous environments, it is still unknown whether a consensus protocol with quadratic communication complexity can be obtained in partial synchrony where the network alternates between (1) asynchronous periods, with unbounded message delays, and (2) synchronous periods, with <span>\\(\\delta \\)</span>-bounded message delays. Until now, the most efficient known solutions for Byzantine consensus in partially synchronous settings had cubic communication complexity (e.g., HotStuff, binary DBFT). This paper closes the existing gap by introducing <span>SQuad</span>, a partially synchronous Byzantine consensus protocol with <span>\\(O\\big (n^2\\big )\\)</span> worst-case communication complexity. In addition, <span>SQuad</span> is optimally-resilient (tolerating up to <span>\\(f < n / 3\\)</span> failures) and achieves <span>\\(O(f \\cdot \\delta )\\)</span> worst-case latency complexity. The key technical contribution underlying <span>SQuad</span> lies in the way we solve <i>view synchronization</i>, the problem of bringing all correct processes to the same view with a correct leader for sufficiently long. Concretely, we present <span>RareSync</span>, a view synchronization protocol with <span>\\(O\\big (n^2\\big )\\)</span> communication complexity and <span>\\(O(f \\cdot \\delta )\\)</span> latency complexity, which we utilize in order to obtain <span>SQuad</span>.</p>","PeriodicalId":50569,"journal":{"name":"Distributed Computing","volume":"30 26 1","pages":""},"PeriodicalIF":1.3000,"publicationDate":"2023-12-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Byzantine consensus is $$\\\\Theta (n^2)$$ : the Dolev-Reischuk bound is tight even in partial synchrony!\",\"authors\":\"Pierre Civit, Muhammad Ayaz Dzulfikar, Seth Gilbert, Vincent Gramoli, Rachid Guerraoui, Jovan Komatovic, Manuel Vidigueira\",\"doi\":\"10.1007/s00446-023-00458-w\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"<p>The Dolev-Reischuk bound says that any deterministic Byzantine consensus protocol has (at least) quadratic (in the number of processes) communication complexity in the worst case: given a system with <i>n</i> processes and at most <span>\\\\(f < n / 3\\\\)</span> failures, any solution to Byzantine consensus exchanges <span>\\\\(\\\\Omega \\\\big (n^2\\\\big )\\\\)</span> words, where a word contains a constant number of values and signatures. While it has been shown that the bound is tight in synchronous environments, it is still unknown whether a consensus protocol with quadratic communication complexity can be obtained in partial synchrony where the network alternates between (1) asynchronous periods, with unbounded message delays, and (2) synchronous periods, with <span>\\\\(\\\\delta \\\\)</span>-bounded message delays. Until now, the most efficient known solutions for Byzantine consensus in partially synchronous settings had cubic communication complexity (e.g., HotStuff, binary DBFT). This paper closes the existing gap by introducing <span>SQuad</span>, a partially synchronous Byzantine consensus protocol with <span>\\\\(O\\\\big (n^2\\\\big )\\\\)</span> worst-case communication complexity. In addition, <span>SQuad</span> is optimally-resilient (tolerating up to <span>\\\\(f < n / 3\\\\)</span> failures) and achieves <span>\\\\(O(f \\\\cdot \\\\delta )\\\\)</span> worst-case latency complexity. The key technical contribution underlying <span>SQuad</span> lies in the way we solve <i>view synchronization</i>, the problem of bringing all correct processes to the same view with a correct leader for sufficiently long. Concretely, we present <span>RareSync</span>, a view synchronization protocol with <span>\\\\(O\\\\big (n^2\\\\big )\\\\)</span> communication complexity and <span>\\\\(O(f \\\\cdot \\\\delta )\\\\)</span> latency complexity, which we utilize in order to obtain <span>SQuad</span>.</p>\",\"PeriodicalId\":50569,\"journal\":{\"name\":\"Distributed Computing\",\"volume\":\"30 26 1\",\"pages\":\"\"},\"PeriodicalIF\":1.3000,\"publicationDate\":\"2023-12-11\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Distributed Computing\",\"FirstCategoryId\":\"94\",\"ListUrlMain\":\"https://doi.org/10.1007/s00446-023-00458-w\",\"RegionNum\":4,\"RegionCategory\":\"计算机科学\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q3\",\"JCRName\":\"COMPUTER SCIENCE, THEORY & METHODS\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Distributed Computing","FirstCategoryId":"94","ListUrlMain":"https://doi.org/10.1007/s00446-023-00458-w","RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q3","JCRName":"COMPUTER SCIENCE, THEORY & METHODS","Score":null,"Total":0}
Byzantine consensus is $$\Theta (n^2)$$ : the Dolev-Reischuk bound is tight even in partial synchrony!
The Dolev-Reischuk bound says that any deterministic Byzantine consensus protocol has (at least) quadratic (in the number of processes) communication complexity in the worst case: given a system with n processes and at most \(f < n / 3\) failures, any solution to Byzantine consensus exchanges \(\Omega \big (n^2\big )\) words, where a word contains a constant number of values and signatures. While it has been shown that the bound is tight in synchronous environments, it is still unknown whether a consensus protocol with quadratic communication complexity can be obtained in partial synchrony where the network alternates between (1) asynchronous periods, with unbounded message delays, and (2) synchronous periods, with \(\delta \)-bounded message delays. Until now, the most efficient known solutions for Byzantine consensus in partially synchronous settings had cubic communication complexity (e.g., HotStuff, binary DBFT). This paper closes the existing gap by introducing SQuad, a partially synchronous Byzantine consensus protocol with \(O\big (n^2\big )\) worst-case communication complexity. In addition, SQuad is optimally-resilient (tolerating up to \(f < n / 3\) failures) and achieves \(O(f \cdot \delta )\) worst-case latency complexity. The key technical contribution underlying SQuad lies in the way we solve view synchronization, the problem of bringing all correct processes to the same view with a correct leader for sufficiently long. Concretely, we present RareSync, a view synchronization protocol with \(O\big (n^2\big )\) communication complexity and \(O(f \cdot \delta )\) latency complexity, which we utilize in order to obtain SQuad.
期刊介绍:
The international journal Distributed Computing provides a forum for original and significant contributions to the theory, design, specification and implementation of distributed systems.
Topics covered by the journal include but are not limited to:
design and analysis of distributed algorithms;
multiprocessor and multi-core architectures and algorithms;
synchronization protocols and concurrent programming;
distributed operating systems and middleware;
fault-tolerance, reliability and availability;
architectures and protocols for communication networks and peer-to-peer systems;
security in distributed computing, cryptographic protocols;
mobile, sensor, and ad hoc networks;
internet applications;
concurrency theory;
specification, semantics, verification, and testing of distributed systems.
In general, only original papers will be considered. By virtue of submitting a manuscript to the journal, the authors attest that it has not been published or submitted simultaneously for publication elsewhere. However, papers previously presented in conference proceedings may be submitted in enhanced form. If a paper has appeared previously, in any form, the authors must clearly indicate this and provide an account of the differences between the previously appeared form and the submission.