A method for evaluating uncertainties in the early development phases of embedded real-time systems

In the early phases of embedded real-time systems development, design decisions are made with large consequences on the quality of the resulting product. At the same time, only limited information is available about key attributes, which makes it difficult to get correct decisions. In this paper, it is studied how to include uncertainty in the information in early phases and to handle the risks involved. Using Monte Carlo simulation, the uncertainties are evaluated based on estimations of basic parameters. The technique is applied to schedulability analysis, and it is shown that the results can be counter-intuitive, and thus that current design practice can lead to higher risk than assumed by the designers.