The purpose of this paper is to present a new algorithm for distributed event-driven simulation. The algorithm is illustrated using two examples: a token transport network and an asynchronous Ising system. The algorithm is a combination of three elements: (1) the bounded lag restriction; (2) minimum propagation delays; and (3) opaque periods. The bounded lag restriction refers to timing of events as they arrive in the simulation. Opaque periods refers to “the ‘promise’ of a node not to transfer information along a certain outgoing link for a certain period of simulated time.”
The paper includes benchmarks obtained from executing the algorithm on both a Sequent’s Balance (multiple-instruction multiple–data type machine) and a Connection Machine (single-instruction multiple–data type machine). Its only drawback is the author’s assumption that the reader will be familiar with the two systems used as examples; otherwise, the paper is well written and complete, and it contributes to the subject of distributed event-driven simulation.