Integrating more than one hard real-time devices while making sure real-time demands is an intricate task. Model-based approaches are available for aiding developers with validating real-time requirements. Hereby, a developer identifies (expected) execution timings, like the Worst Case Execution Time (WCET). These models function as input for model checkers designed to use the specified timings to validate the real-time requirements. No methods/approaches to test and ensure the correctness of the specified timings were found. This study is designed to close this gap. Initially, a new Domain-Specific Language (DSL) was created to model the system. This had great similarity with Architecture Analysis & Design Language (AADL). Because AADL is definitely an industry standard and existing tools could be used again, AADL was selected to model the system to be developed. AADL is extended with the ability to specify monitors, which monitor execution scenarios….
Contents
1 Introduction
1.1 Context
1.2 Problem statement
1.3 Approach
1.4 Contributions
1.5 Outline
2 Case study: Medical devices
2.1 The integrated system
2.2 Neuronavigation
2.2.1 VectorVision
2.3 Haptic devices
2.3.1 Passive haptic device
2.3.2 Active haptic device
2.3.3 Freedom6S
2.4 Challenges
2.4.1 Brain shift
2.4.2 Missing a deadline
2.4.3 Frequency mismatch
2.5 Previous work regarding the case study
II Technological background and related work
3 Real-Time Safety-critical Systems
3.1 What is real time
3.2 Safety-critical systems
ii3.3 Performance analysis
3.3.1 Static Analysis
3.3.2 Dynamic Analysis
4 Domain-specific languages
4.1 What is a domain specific language?
4.2 External & Internal DSLs
4.3 Advantages and disadvantages
4.4 When & how to create a DSL
4.5 Which real-time oriented DSLs exists?
4.5.1 Functional DSLs
4.5.2 Model based DSLs
4.5.3 Miscellaneous
4.6 Architecture Analysis & Design Language
4.6.1 Software components
4.6.2 Properties
4.6.3 Analysis of AADL models
III Solution design & implementation
5 DSL design
5.1 Solution technology
5.2 Analysis of the domain
5.2.1 De ning the problem domain
5.2.2 Domain analysis
5.3 Designing the DSL
5.4 Implementing the DSL
5.5 Comparing the DSL with AADL
5.6 Conclusion
6 Integration Framework
6.1 Framework design
6.2 Linking the framework & AADL
6.3 Conclusion
7 Performance Monitors
7.1 Design
7.2 Semantics
7.2.1 Statechart design
7.2.2 Examples
7.3 Implementation decisions
7.3.1 Multiple threads on a single method
7.3.2 How to monitor deadlines
iii7.3.3 Where to determine the time
7.4 Monitors & AADL
7.5 Conclusion
8 Case study: Integrating medical devices
8.1 Design
8.2 Implementation
8.3 Results
8.4 Conclusion
IV Evaluation & Conclusions
9 Evaluation
9.1 Framework
9.1.1 Test set-up
9.1.2 Overhead and scalability tests
9.1.3 Timer jitter
9.1.4 Reusability of the framework
9.1.5 Reusability of formal analysis tools
9.2 Case study
9.2.1 Performance evaluation
9.2.2 Prerequisites for real-world usage
9.3 Conclusion
10 Conclusions
10.1 Answers to research questions
11 Recommendations & future work
Bibliography
Source: University of Twente
Download URL 2: Visit Now