Architectural Modeling and Analysis of Complex Real-Time Systems

Most automation systems and other large industrial software systems have long lifetimes, and customers expect these systems to be supported as long as they are in operation. Furthermore, software components in these systems may be reused in different products, e.g. using a software product line approach. Hence, the lifetime of software in individual systems may be very long; several decades or even longer.Software that is used for a long time will be exposed to frequent changes as the system evolve over time, e.g. due to adding new functionality, error corrections, or changing the hardware platform. The larger and older the system is, the harder it becomes to foresee the consequences of changes.In this thesis we present three different techniques for managing the evolution of large and complex real-time systems. The techniques are based on analytical modeling, predicting different quality properties, e.g. temporal correctness, by analyzing a model of the software. The first technique is a component model with analytical interfaces (ReFlex) that allows us to predict different properties of a component assembly, the second is a probabilistic modeling language which is analyzed by simulations (ART-FW), and the third technique is an extension of classical timed automata with a notion of real-time tasks (TAT).Ideally, the analytical models should evolve together with the software. However, since new features are often added and the implementation is often changed without updating the model, the model becomes obsolete and predictions based on the model are no longer valid. By applying the techniques proposed in this thesis, we can re-introduce analyzability; Using ReFlex we can update the analytical aspects while re-designing the system. Unless ReFlex has been used in the earlier design, this will require a costly redesign of the complete system, but consistency between the analytical model and the implementation will be ensured. Using ART-FW or TAT the implementation will be kept untouched by introducing a separate model. The drawback is that an extra effort is required to keep the model consistent with the implementation.We have applied ART-FW in the re-engineering activity of a large industrial system. The results indicate that the approach is indeed applicable on real systems.