@MISC{Bygde07parametricwcet, author = {Stefan Bygde}, title = {Parametric WCET Analysis}, year = {2007} }

Share

OpenURL

Abstract

WCET analysis is about getting information about the worst-case execution time for a program on a certain platform for any possible input. Static WCET analysis is an automatic way to derive a safe upper bound for the worst-case execution time of a program by analyzing its source code and the hardware of the target platform. In many cases it is useful to have such an estimation of the WCET, especially for hard real-time systems. Normaly, the WCET is estimated by measurements of the executing program (dynamic analysis), which gives an underestimation of the actual WCET (because it is not possible to be sure that the worst case has been measured). In static analysis on the other hand, the estimation of the WCET should always be equal or greater than the actual WCET (ofcourse, assumed that the analysis tool is correct). In classic static WCET analysis, the upper bound of the execution time is given as an absolute value (for instance, in milliseconds) whereas parametric WCET-analysis aims to derive the WCET-bound as an expression of some undetermined variables, which later can be instantiated.