Communication Pattern Models: An Extension of Action Models for Dynamic-Network Distributed Systems (2106.11508v1)

Abstract: Halpern and Moses were the first to recognize, in 1984, the importance of a formal treatment of knowledge in distributed computing. Many works in distributed computing, however, still employ informal notions of knowledge. Hence, it is critical to further study such formalizations. Action models, a significant approach to modeling dynamic epistemic logic, have only recently been applied to distributed computing, for instance, by Goubault, Ledent, and Rajsbaum. Using action models for analyzing distributed-computing environments, as proposed by these authors, has drawbacks, however. In particular, a direct use of action models may cause such models to grow exponentially as the computation of the distributed system evolves. Hence, our motivation is finding compact action models for distributed systems. We introduce communication pattern models as an extension of both ordinary action models and their update operator. We give a systematic construction of communication pattern models for a large variety of distributed-computing models called dynamic-network models. For a proper subclass of dynamic-network models called oblivious, the communication pattern model remains the same throughout the computation.