Prozessorientierte optimistisch-parallele Simulation

The main motivation of parallel discrete event-driven simulations (PDES) is a speedup of simulation runs. This aim is reflected by two similarities shared by almost all current PDES implementations: For efficiency reasons the event-oriented modeling view is used exclusively and the implementation is done in a low-level programming language. However, both properties impose serious constraints on the simulation model. In many cases the event-oriented modeling view – especially in contrast to the processoriented view – prevents a structurally equivalent model of the original system. Furthermore, the actual implementation often leads to a further decomposition of the model, due to a lack of abstract concepts in the low-level programming language. The resulting implementations run efficiently, but at the price of a low comprehensibility, due to the structural difference from the original problem. Consequently, the efforts for validation or modification of the simulation model are considerably increased. These disadvantages led to a different development route in the sequential simulation community – in which the execution time is of minor importance. Here, the combined use of the event-oriented and the process-oriented view is common practice in order to obtain a structurally equivalent and intuitive model. Moreover, the simulation model is usually implemented in a higher-level programming language, e.g. Java. A combination of the advantages of the sequential, as well as the parallel simulation approach is desirable, especially in the case of large and complex simulation scenarios. These are only manageable if a corresponding structurally equivalent simulation model is employed. Also, the very same models will profit most from a speed-up by parallelization. Unfortunately, the implementation of a process-oriented view in a parallel fashion is not a trivial task. Parallel simulation implementations generally suffer from additional computational cost that can offset the parallel speed-up and even lead to