Regression Analysis-Based Cost Estimation of Simulation Models in Model Cost Database*

The effectiveness of simulation-based acquisition, which is a robust, collaborative use of modeling and simulation (M&S) technology throughout s acquisition phases and programs, should be estimated quantifiably, even in a rough order of magnitude manner, with the consideration of M&S overhead. This paper presents a simulation model cost database that can estimate the development costs of simulation models based on regression analysis. When simulation model that have been newly developed throughout acquisition programs are inserted into the databased, size-to-cost trends are automatically computed for all data sets in the database. Based on these trends, the development costs of simulation models to be developed can be estimated if their sizes can be roughly estimated. The database may be used in parts of the effectiveness estimation of the use of modeling and simulation technologies in acquisition programs. Moreover, it may be used to verify various methods of simulation model size measurement and find out the most promising ones. We have implemented a prototype of the database and are working on the implementation of a regression analysis based estimation module.