Leaderless Byzantine Fault Tolerance

Tian Qin

EECS Department
University of California, Berkeley
Technical Report No. UCB/EECS-2020-121
May 29, 2020

http://www2.eecs.berkeley.edu/Pubs/TechRpts/2020/EECS-2020-121.pdf

In this work, we propose Leaderless Byzantine Fault Tolerance (LBFT), a novel consensus algorithm that combines Snowball and Practical Byzantine Fault Tolerance (pBFT), two existing consensus algorithms. Achieving consensus in decentralized systems has been difficult as they lack certain properties that many algorithms assume. Our approach looks to take advantage of the decentralized aspect of Snowball and the deterministic property of pBFT so that the weaknesses of Snowball’s probabilistic property and pBFT’s reliance on "leader" are eliminated. The algorithm we propose is applicable to decentralized systems such as blockchain systems even though Snowball and pBFT both make stronger assumptions than decentralized systems allow. By simulating real-world environments and comparing with pBFT performances, we show that LBFT is feasible in the context of real-world decentralized systems and has sufficient performances that are close to those of pBFT.

Advisor: Dawn Song


BibTeX citation:

@mastersthesis{Qin:EECS-2020-121,
    Author = {Qin, Tian},
    Title = {Leaderless Byzantine Fault Tolerance},
    School = {EECS Department, University of California, Berkeley},
    Year = {2020},
    Month = {May},
    URL = {http://www2.eecs.berkeley.edu/Pubs/TechRpts/2020/EECS-2020-121.html},
    Number = {UCB/EECS-2020-121},
    Abstract = {In this work, we propose Leaderless Byzantine Fault Tolerance (LBFT), a novel consensus algorithm that combines Snowball and Practical Byzantine Fault Tolerance (pBFT), two existing consensus algorithms. Achieving consensus in decentralized systems has been difficult as they lack certain properties that many algorithms assume. Our approach looks to take advantage of the decentralized aspect of Snowball and the deterministic property of pBFT so that the weaknesses of Snowball’s probabilistic property and pBFT’s reliance on "leader" are eliminated. The algorithm we propose is applicable to decentralized systems such as blockchain systems even though Snowball and pBFT both make stronger assumptions than decentralized systems allow. By simulating real-world environments and comparing with pBFT performances, we show that LBFT is feasible in the context of real-world decentralized systems and has sufficient performances that are close to those of pBFT.}
}

EndNote citation:

%0 Thesis
%A Qin, Tian
%T Leaderless Byzantine Fault Tolerance
%I EECS Department, University of California, Berkeley
%D 2020
%8 May 29
%@ UCB/EECS-2020-121
%U http://www2.eecs.berkeley.edu/Pubs/TechRpts/2020/EECS-2020-121.html
%F Qin:EECS-2020-121