Timing and scheduling analysis of real-time object-oriented models

The increasing complexity of real-time software has led to a recent trend in the use of high-level modelling languages for the development of real-time software. One representative example is the Real-Time Object Oriented Modeling (ROOM) language, which provides features such as encapsulation, polymorphism, inheritance, state machine descriptions of system behavior, formal semantics for executability of models and the possibility of automated code generation. The full benefits of the ROOM language are obtained through the use of the ObjecTime toolset, designed to support the ROOM language and its development process in order to automatically create an executable for a target platform equipped with a real-time operating system. However, the ROOM language and the ObjecTime toolset largely ignore the temporal aspects of real-time systems, and fail to provide any guidance to the designer about predicting and analyzing the temporal behavior of their intended applications. The main objective of this thesis is to develop ways to perform such timing and scheduling analysis for single- and multi-threaded ROOM models. This work builds on results presented in (SFR97) and (SPFR98), where guidelines for the design and implementation of real-time object oriented (ROOM) models were developed and tested. In this thesis, we shall consider three orthogonal timing analysis methods: real-time scheduling theory, symbolic model checking and discreet task simulation. We formulate design guidelines for single-and multi-threaded ROOM executables. Then we develop and validate canonical scheduling models for each type of executable. Finally, we present a case study of an automobile cruise control system to further illustrate the concepts presented in this thesi