Title: On Probabilistic Byzantine Fault Tolerance

Abstract: Byzantine fault tolerance (BFT) has been extensively studied in distributed trustless systems to guarantee system's functioning when up to 1/3 Byzantine processes exist. Despite a plethora of previous work in BFT systems, they are mainly concerned about common knowledge deducible from the states of all participant processes. In BFT systems, it is crucial to know about which knowledge a process knows about the states of other processes and the global state of the system. However, there is a lack of studies about common knowledge of Byzantine faults, such as, whether existence of a Byzantine node is known by all honest nodes. In a dynamic setting, processes may fail or get compromised unexpectedly and unpredictably. It is critical to reason about which processes know about the faulty processes of the network.
In this paper, we are interested in studying BFT systems in which Byzantine processes may misbehave randomly, possibly at some random periods of time. The problem of \emph{probabilistic Byzantine} (PB) processes studied in this paper is more general than the problems previously studied in existing work. We propose an approach that allows us to formulate and reason about the concurrent knowledge of the PB processes by all processes. We present our study of the proposed approach in both synchronous and asynchronous systems.