| Abstrakt: |
In this chapter we want to substantiate our previous claim that the system architecture of a target system plays a major role in developing a test methodology for this system. So far, the term "distributed system" was used only in its intuitive meaning; we now give a more rigorous characterization of the system model. We also introduce the notions of event-triggered and time-triggered systems, a distinction which we have found to be very useful in our work on real-time systems [Kop90c, Kop91a]. We then undertake a comparison of these two types of systems, with the result that time-triggered systems offer some significant advantages with respect to testing. The criteria affected by this comparison are Observability, Controllability, and the Test Coverage. [ABSTRACT FROM AUTHOR] |
