Integrating Concurrent and Conservative Distributed Discrete-Event Simulators

There have been many attempts to apply parallel computers to discrete-event simulation. These may be divided into two main approaches, distributed simulation and concurrent simulation. Distributed simulation relies on a spatial decomposition and partitions the simulation model into components that can be executed on different processors. Concurrent simulation is based on a temporal decomposition. This paper focuses on the integration of these two approaches; we call our result an integrated conservative distributed concurrent simulator. This is a distributed simulator with concurrency added to each model component. To do so, a shared- memory environment was used, and both approaches were unified to an event- centered view. We also partitioned the global concurrent pending-event set of the concurrent simulator and provided each model component in the distributed simulator with a local concurrent pending-event set, allowing us to add concurrency to each model component.