Cornell University
Library
Cornell UniversityLibrary

eCommons

Help
Log In(current)
  1. Home
  2. Cornell Computing and Information Science
  3. Computer Science
  4. Computer Science Technical Reports
  5. Asynchronous Consensus and Byzantine Protocols in Faulty Environments

Asynchronous Consensus and Byzantine Protocols in Faulty Environments

File(s)
83-559.ps (409.13 KB)
83-559.pdf (1.97 MB)
Permanent Link(s)
https://hdl.handle.net/1813/6399
Collections
Computer Science Technical Reports
Author
Bracha, Gabriel
Toueg, Sam
Abstract

A consensus protocol enables a system of $n$ aynchronous processes, some of which are faulty, to reach agreement. There are two kinds of faulty processes: fail-stop processes can only die, malicious processes can also send false messages. We investigate consensus protocols that terminate within finite time with probability 1 under certain assumptions on the behavior of the system. With fail-stop processes, we show that $\lceil (n + 1)/2 \rceil$ correct processes are necessary and sufficient to reach agreement. In the malicious case, we show that $\lceil (2n + 1)/3 \rceil$ correct processes are necessary and sufficient to reach agreement. This is contrasted with a recent result that there is no consensus protocol for the fail-stop case that always terminates within a bounded number of steps, even if only one process can fail. We also investigate the possibility of reliable broadcast (Byzantine Agreement) in an asynchronous system. We define the notion of Asynchronous Byzantine Agreement, and show that $\lceil (2n + 1)/3 \rceil$ correct processes are necessary and sufficient to reach Asynchronous Byzantine Agreement.

Date Issued
1983-06
Publisher
Cornell University
Keywords
computer science
•
technical report
Previously Published as
http://techreports.library.cornell.edu:8081/Dienst/UI/1.0/Display/cul.cs/TR83-559
Type
technical report

Site Statistics | Help

About eCommons | Policies | Terms of use | Contact Us

copyright © 2002-2026 Cornell University Library | Privacy | Web Accessibility Assistance