| Parallel and distributed simulation is recognized as a viable method for the modeling of complex dynamic systems. The main benefits of using a parallel versus a sequential technique in solving these problems are decreasing execution time and distributing the memory used in the simulation over a number of processors.;Parallel simulation techniques have successfully been applied to the modeling of cellular communication services. They focus on defining efficient models for simulation, on the study of protocols or on the performance analysis of resource allocation algorithms. In this thesis, we continue the research in this domain by defining a model focusing on simulation accuracy. We will also test the limits on the size of the application model that can be simulated. Two channel allocation techniques are implemented. The first is based on fixed channel allocation and the second on a technique known as channel segregation. The latter technique requires that interference data be computed. For this purpose, we describe a brute force interference calculation algorithm. (Abstract shortened by UMI.)... |
