DROPS (Schloss Dagstuhl – Leibniz Center for Informatics) · 2019 · 17 citations · 4 references
ComputingComputer SystemsByzantine StudiesDecentralizationDistributed SystemsConsensus AlgorithmsConsensus ProtocolsDecentralized ControlDistributed AlgorithmsDistributed ConsensusBroadcastByzantine BroadcastFederated ConsensusOpen MembershipQuorum SystemsQuorum-based CoordinationQuorum-based ProtocolsQuorum-based Systems
Some of the recent blockchain proposals, such as Stellar and Ripple, use quorum-like structures typical for Byzantine consensus while allowing for open membership. This is achieved by constructing quorums in a decentralised way: each participant independently chooses whom to trust, and quorums arise from these individual decisions. Unfortunately, the theoretical foundations underlying such blockchains have not been thoroughly investigated. To close this gap, in this paper we study decentralised quorum construction by means of federated Byzantine quorum systems, used by Stellar. We rigorously prove the correctness of basic broadcast abstractions over federated quorum systems and establish their relationship to the classical Byzantine quorum systems. In particular, we prove correctness in the realistic setting where Byzantine nodes may lie about their trust choices. We show that this setting leads to a novel variant of Byzantine quorum systems where different nodes may have different understanding of what constitutes a quorum.
4
Ramakrishna Kotla, Lorenzo Alvisi, Mike Dahlin et al. · 2007 · 642 citations
Dahlia Malkhi, Michael K. Reiter · Distributed Computing · 1998 · 527 citations
Making Byzantine fault tolerant systems tolerate Byzantine faults
Allen Clement, Edmund Wong, Lorenzo Alvisi et al. · 2009 · 367 citations · Full text