An approach to object-oriented modeling and performance evaluation

Abstract

Diverse modeling and simulation tools are applied in extensive applications. The approaches used range from programming with a specific simulation description language to automation using an icon-driven user interface. The advantage in utilizing simulation is to assess the system's performance prior to an actual implementation. Functionality, maintainability, and expansibility are the primary criteria used to make a choice of a specific tool. To strengthen these criteria, a general-purpose environment called performance object-oriented modeling and simulation environment (POSE) has been developed. The objectives of POSE are to automatically construct simulation models for the systems to be designed, to efficiently define the system performance measures, and to accurately generate the performance data expected. The environment is briefly summarized and an application study for a multiprocessor computer system is presented.<>