Oana Florescu

Software/Hardware Engineering with the Parallel Object-Oriented Specification Language

The complexity of designing hardware/software systems motivates research on frameworks that structure and automate the design process. Such design methodologies reduce the risk of expensive design-implementation iterations by assisting designers in constructing models. Software/hardware engineering (SHE) is a general-purpose system-level design methodology that supports analysing both functional correctness and performance properties. SHE combines the Unified Modelling Language with the parallel object-oriented specification language to specify models. The designer is assisted in constructing models using these languages and applying the analysis techniques with various guidelines and modelling patterns. A key feature of SHE is its foundation on formal methods, which ensures that the obtained analysis results are unambiguous. SHE also includes guidelines and techniques for automatic synthesis of real-time control software. This is again based on formal methods to ensure that properties in a model (including real-time properties) are preserved by the software realisation. Finally, to enable an effective and efficient application of the modelling languages as well as the analysis and synthesis techniques, SHE is accompanied with a set of user-friendly tools. This paper gives an overview of SHE, thereby briefly touching upon the underlying mathematical foundation of the analysis and synthesis techniques as well as upon some open issues that require further research.

Predictability in Real-Time System Development

The large gap existing between requirements and realizations has been a pertinacious problem in complex system design. This holds in particular for realtime systems with strict timing constraints and critical-safety requirements. Designers have to rely on a multi-step design process, where design decisions are made at different modelling levels. To ensure the effectiveness of this design process, predictability should be well-supported by design approaches, allowing designers to predict properties of future design outcomes based on existing design results. In this chapter, we first discuss the role of the semantics of design languages and investigated how they can support a predictable design process. Then, the deficiencies, w.r.t. predictability support, of existing design approaches for real-time systems are illustrated by an example. Finally, a predictable design approach for real-time systems is introduced to overcome this problem.

Patterns for Automatic Generation of Soft Real-time System Models

Worst-case assumptions about the timing of systems are often too conservative when analyzing distributed soft real-time systems as they lead to over-dimensioned and expensive products. For these systems, a certain percentage of deadline misses is often affordable. Instead of a binary answer regarding the schedulability of such a system, a more interesting metric is the degree to which the system meets the timing requirements. For this, an appropriate model that realistically expresses the behavior of a soft real-time system when deployed on a multiprocessor platform should be built and analyzed. In this article, we present such a modeling approach based on the formal modeling language POOSL (parallel object-oriented specification language). Moreover, to alleviate the process of modeling, we present a pattern-based description language that allows an application, together with the multiprocessor platform and the deployment to be described in a concise way. Such a pattern-based description can be translated automatically into an executable POOSL model through which performance properties can be analyzed based on simulations. The suitability of our approach is demonstrated by exploring the design space of a distributed in-car radio navigation system.

Error computation for predictable real-time software synthesis

Synthesizing an implementation from a model in a predictable way is one of the major challenges in real-time system design. In our previous work we addressed this problem by generating in real-time an execution path through a model and by synchronizing the model time with the physical time. The execution path as observed in model time has a time difference with the execution path as observed in physical time. This distance determines the extent to which real-time model properties are preserved in the implementation. The key contribution of this article is an analytical approach for calculating the distance between a model and a corresponding implementation. Based on this distance, the real-time properties of the implementation can be predicted from the model. A paper path of a printer is used as a case study to show the effectiveness of the technique.

Reusing systems design experience through modelling patterns

Based on design experience forreal-time systems, we introduce modelling patterns to enable easy composition of models for design space exploration. Our proposed approach does not require deep knowledge of the modelling language used for the actual specification of the model and its related analysis techniques. The patterns proposed in this paper cover different types of real-time tasks, resources and mappings, and include also aspects that are usually ignored in classical analysis approaches, like task activation latency or execution context switches. In this paper, we present a library of such modelling patterns expressed in the POOSL language. By identifying the patterns that are needed for the specification of a system, the POOSL model can be automatically generated and analysed using its related tools and techniques as illustrated in two case studies.

Property-Preservation Synthesis for Unified Control- and Data-Oriented Models

Process for recovering water of reduced salinity by contacting a semipermeable hydrophilic composite sheet with high salinity water and recovering a lower salinity water from the water sorbed in the sheet by applying sufficient physical pressure to force the water out of the sheet.