Self-Stabilizing Byzantine Token Circulation

There is an abundance of writing about Token Circulation (or leader election). Much of the work is dedicated to self-stabilizing Token Circulation, ever since the publishing of Dijkstra’s seminal paper. Few of the papers focus on the Byzantine fault model and, to the best of our knowledge, there is no self-stabilizing Token Circulation algorithm that tolerates Byzantine faults. In this paper, we present an elegant selfstabilizing Byzantine Token Circulation algorithm that has comparably good time and message complexities. It has optimal fairness and every node holds the token 1/n part of the time in the long run, where n is the number of nodes in the network. Our protocol is based on a tight Byzantine self-stabilizing pulse synchronization procedure. The synchronized pulses are used as events for initializing Byzantine consensus on the id of the next node to hold the token. When the system performs well the time complexity of our scheme is minimal, merely two communication rounds. When the system converges to a desired state following a chaotic situation the additional overhead is that of Byzantine consensus (O(f ′), where f ′ is the actual number of faulty nodes).