This paper introduces regular extrapolation, a technique that provides descriptions of systems or system aspects a posteriori in a largely automatic way. The descriptions come in the form of models which offer the possibility of mechanically producing system tests, grading test suites and monitoring running systems. Regular extrapolation builds models from observations via techniques from machine learning and finite automata theory. Also expert knowledge about the system enters the model construction in a systematic way. The power of this approach is illustrated in the context of a test environment for telecommunication systems.
Automatically generated models may provide the key towards controlling the evolution of complex systems, form the basis for test generation and may be applied as monitors for running applications. However, the practicality of automata learning is currently largely preempted by its extremely high complexity and unrealistic frame conditions. By optimizing a standard learning method according to domainspecific structural properties, we are able to generate abstract models for complex reactive systems. The experiments conducted using an industrylevel test environment on a recent version of a telephone switch illustrate the drastic effect of our optimizations on the learning efficiency. From a conceptual point of view, the developments can be seen as an instance of optimizing general learning procedures by capitalizing on specific application profiles.
We elaborate on the theoretical foundation and practical application of the contract-based specification method originally developed in the Integrated Project SPEEDS [11], [9] for two key use cases in embedded systems design. We demonstrate how formal contract-based component specifications for functional, safety, and real-time aspects of components can be expressed using the pattern-based requirement specification language RSL developed in the Artemis Project CESAR, and develop a formal approach for virtual integration testing of composed systems based on such contract-specifications of subsystems. We then present a methodology for multi-criteria architecture evaluation developed in the German Innovation Alliance SPES on Embedded Systems.
Abstract. We propose algorithms significantly extending the limits for maintaining exact representations in the verification of linear hybrid systems with large discrete state spaces. We use AND-Inverter Graphs (AIGs) extended with linear constraints (LinAIGs) as symbolic representation of the hybrid state space, and show how methods for maintaining compactness of AIGs can be lifted to support model-checking of linear hybrid systems with large discrete state spaces. This builds on a novel approach for eliminating sets of redundant constraints in such rich hybrid state representations by a suitable exploitation of the capabilities of SMT solvers, which is of independent value beyond the application context studied in this paper. We used a benchmark derived from an Airbus flap control system (containing 2 20 discrete states) to demonstrate the relevance of the approach.
Abstract.We address the problem of model checking hybrid systems which exhibit nontrivial discrete behavior and thus cannot be treated by considering the discrete states one by one, as most currently available verification tools do. Our procedure relies on a deep integration of several techniques and tools. An extension of AND-Inverter-Graphs (AIGs) with first-order constraints serves as a compact representation format for sets of configurations which are composed of continuous regions and discrete states. Boolean reasoning on the AIGs is complemented by firstorder reasoning in various forms and on various levels. These include implication checks for simple constraints, test vector generation for fast inequality checks of boolean combinations of constraints, and an exact subsumption check for representations of two configurations.These techniques are integrated within a model checker for universal CTL. Technically, it deals with discrete-time hybrid systems with linear differentials. The paper presents the approach, its prototype implementation, and first experimental data.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.