An application of complex task modeling

The paper presents some experimental results from scheduling analysis of a task set on a commercial real-time operating system using a complex task modeling and analysis framework. Simple task models, used in traditional scheduling analysis techniques, are only concerned with a task's execution time. Complex task models examine shared resource contention and varying priorities, resulting in an enhanced view of a task's behavior. A series of experiments using VxWorks(R) generates complex task models for the tasks executing on a system, predicts system behavior, and explores the limits of current modeling and analysis techniques.