Title:Fast Internet Computer Consensus

Abstract:This paper presents the first rotating leader state machine replication (SMR) protocol that allows transactions to be confirmed in just a single round-trip time in the Byzantine fault tolerance (BFT) setting. Based on minimal alterations to the Internet Computer Consensus (ICC) protocol and with negligible communication overhead, we introduce a novel dual mode mechanism that enables optimal block finalization latency in the fast path. Crucially, the modes of operation are integrated, such that even if the fast path is not effective, no penalties are incurred. Moreover, our algorithm maintains the core attributes of the original ICC protocol, including optimistic responsiveness and rotating leaders without the necessity for a view-change protocol.
We prove the correctness of our Fast Internet Computer Consensus (FICC) protocol and provide an open-source implementation of it. Both the FICC and original ICC protocol are compared in a globally distributed wide-area network. Our evaluation reveals that the FICC protocol achieves reduced latency compared to the ICC protocol, without requiring additional security assumptions. Furthermore, by increasing the number of replicas to $n = 5f + 1$, we exhibit that latency improvements close to the theoretical maximum of 33% are attainable. We conclude by highlighting the network topology as a significant factor in evaluating and comparing the latency of consensus algorithms.