Abstract
Rate monotonic scheduling provides analytic techniques that allow the feasibility and processor utilization of a set of periodically scheduled tasks to be computed. However, the scheduling model must be adapted to take into account specific implementation details. Moreover, a model is not the same thing as the actual system; the accuracy and reliability of a model must be verified. This paper presents an Ada coding template for rate monotonic scheduling and models for that template. A series of experiments is described whose purpose was to compare the analytic predictions and measured performance for some sample task sets. Predicted and measured values differed by less than about 3% in all cases.

This publication has 0 references indexed in Scilit: