Skip to main navigation Skip to search Skip to main content

Nova: A robustness-oriented Byzantine Fault Tolerance protocol

  • Xiang Pei*
  • , Yongjian Wang
  • , Zhongzhi Luan
  • *Corresponding author for this work
  • Beihang University

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

With the increased complexity, malicious faults have become an important reasons that affect the reliability of the distributed system, especially the web-scale infrastructures, i.e. Amazon S3, Google AppEngine etc. Most such systems assume benign fault model which can't depict the malicious actions. The goal of Byzantine Fault Tolerance protocol (BFT for short) is to mask the malicious behaviors and it has been proved that some new proposed BFTs are suitable to support practical applications. But these BFTs still lack in robustness, a simple fault injection may cause significantly decrease in throughput or run in low throughput without violating the BFT safety property. We propose a new robustness-oriented BFT named Nova. Experiments show Nova has comparable throughput as PBFT in normal case and behave stably under the malicious attack. Compared with other BFTs, Nova can support practical applications more effectively.

Original languageEnglish
Title of host publicationProceedings - 9th International Conference on Grid and Cloud Computing, GCC 2010
Pages151-156
Number of pages6
DOIs
StatePublished - 2010
Event9th International Conference on Grid and Cloud Computing, GCC 2010 - Nanjing, Jiangsu, China
Duration: 1 Nov 20105 Nov 2010

Publication series

NameProceedings - 9th International Conference on Grid and Cloud Computing, GCC 2010

Conference

Conference9th International Conference on Grid and Cloud Computing, GCC 2010
Country/TerritoryChina
CityNanjing, Jiangsu
Period1/11/105/11/10

Keywords

  • Byzantine Fault Tolerance
  • Equivocation
  • Robustness

Fingerprint

Dive into the research topics of 'Nova: A robustness-oriented Byzantine Fault Tolerance protocol'. Together they form a unique fingerprint.

Cite this