HRT-HOOD(TM): A Structured Design Method for Hard Real-Time Ada Systems

The increasing use of computers for real-time control on board spacecrafts has brought with it a greater emphasis on the development methodology used for such systems. By their nature, spacecraft control computers have to operate unattended for long periods and because of the programmatics of space, systems are subject to a long development cycle. As a result, there are two distinct concerns, the first being that the development approach guarantees functional and timing correctness, the second being that problems, particularly those associated with timing, are considered as early as possible in the spacecraft development life cycle.The European Space Agency has, for a number of years, encouraged the development of software using HOOD. It was thus a natural next step to investigate the incorporation of time within the existing HOOD framework. This has proven to be very beneficial and this book describes the approach developed by the authors for handling Hard Real-Time applications. It describes both the background scheduling theory, provides practical examples of its application to real life problems, and demonstrates how it is used in the various phases of the development of Hard Real-Time systems.

Inhalt

Contents. Foreword. Preface. Acknowledgements. Real-Time Systems Research at York. Part 1: Hard Real-Time HOOD. Overview of the HRT-HOOD design process. Introduction. The importance of non-functional requirements. The software development life cycle. Summary. Logical and physical architecture design in HRT-HOOD. Logical architecture design. Physical architecture design. Summary. HRT-HOOD objects. Graphical representation. Passive objects. Active objects. Protected objects. Cyclic objects. Sporadic objects. Real-time object attributes. The use relationship (control flow). The include relationship (decomposition). Operation decomposition. Object control structure and thread decomposition. Data flows. Exception flows. Environment objects. Class objects. Distributed systems. Summary. Part 2: Mapping HRT-HOOD Designs to Ada. Supporting hard real-time systems in Ada 83 and Ada 95. The Ada 83 and Ada 95 real-time models. Supporting Ada 95 abstractions in Ada 83. Extending the model. Implementation cost. Summary. Overall mapping approach. HOOD 3.1 to Ada 83 mapping. An alternative translation approach. Mapping HRT-HOOD to Ada. Mapping of passive and active objects. Passive terminal objects. Active terminal objects. Class and instance terminal objects. Mapping protected, cyclic and sporadic objects. Protected terminal objects. Cyclic terminal objects. Sporadic terminal objects. Distributed systems. Analysable communication subsystem. Mapping to Ada 95. Mapping protected objects in a distributed Ada environment. Part 3: Case Studies. The Mine Control System. Mine Control System overview. The logical architecture design. The physical architecture design. The Object Description Skeleton (ODS). Translation to Ada 95. Conclusion. The Olympus Attitude and Orbital Control System (AOCS). Background to the case study. The modelled system: the Olympus AOCS. The software architecture design. The physical architecture design. Problems encountered. Summary. Conclusions. Appendices. A: Teminology. B: HRT-HOOD definition rules. Design checking, scoping and HRT-Hood rules. General definitions. Use relationship. Include relationships. Operations. Visibility. Consistency. C: Object Description Skeleton (ODS) syntax summary. General declarations. Object ODS structure. The visible part of the ODS. The hidden part of the ODS. Parameters of class objects. D: Textual formalism - the ODS definition. Passive objects. Active objects. Protected objects. Cyclic objects. Sporadic objects. Environment objects. Class objects. Instances of class objects. E: Device control objects in HRT-HOOD. References. Index.