Skip to main content
European Commission logo print header

Sdl software quality validation for embedded systems" (SDL-SAFE)

Deliverables

The SDL-SAFE project originates from the pressing requirement to test, measure and validate the quality and conformance of SDL generated code for critical communications and control systems. The project has reviewed this requirement in the context of the capabilities of Small and Medium sized Enterprises and seeked to provide a reliable and cost-effective solution. The portable software toolset implements the specification from the Requirements Analysis and follows the software development lifecycle of design, implementation and testing. The software tool implementation follows a progressive prototype approach, which includes the following major milestones: - SDL system Analyser; - SDL system Quality Assessor using Metrics Mechanisms; - Metrics configuration tool; - Assessment reporting tool. The following tools have been developed as parts of the SDL-SAFE Toolset: - A Syntax/Structure Analyser and - A Semantic Analyser. The Syntax/Structure Analyser checks that the SDL system satisfies the SDL language and structure. In particular it checks: - All syntax as defined in SDL language; - All rules and structure use; - Informal propositions. A syntax/structure Analyser provides an interface to configure or override automated checking. It can also be used to manually step thought the checks. The Semantic Analyser checks that the SDL system is defined in a complete manner, that it is self-consistent. The Analyser reads in the SDL text. The analysis of a .PR file identifies syntactic errors and allows semantic checking. The evaluation process includes the definition of quality rules for the SDL notation, which influence the quality criteria of McCall’s quality model. These rules can be grouped according to the elements that influence the quality of an SDL system. The grouping of rules are very helpful to the assessment process, since each group of quality rules affects one ore more quality criteria and there is no need to map each rule separately to the criteria.
The ultimate goal of the SDL-SAFE was to develop an efficient solution for SDL systems quality assessment. The quality improvement of the SDL designs the following benefits for the user members: The design, implementation and testing costs can be sufficiently reduced, the accuracy of testing results can be improved, reliable SDL designs that lead to efficient application systems would be produced. Thus, in order to achieve the goal of providing an efficient toolset that would offer to the users all these benefits, extensive studies were performed especially on mature SDL specifications in order to properly: - Identify their merits and; - To verify the assessment techniques applicability to the entire SDL system development process at different phases of their life cycles. The five user members of the consortium evaluated the document and software deliverables by adapting their existing SDL development methodologies to accommodate the tool. PR files from a range of different protocol profiles were used and are presented in the paragraphs of this report, which is the assessment report of the user trial findings and user feedback. The report includes the comments of experienced test engineers on the efficiency of the solutions and whether they confirm or contradict perception of high and low quality PR files. It also includes recommendations, enhancements to SDL system development cycle and guidelines on tool use. The aim of the Trial and Assessment report is to evaluate the methodology document and the produced tools of the SDL-SAFE toolset. The user evaluation results are the outcomes of the SME user field trials of the document and tool to a range of different SDL systems. Additionally, the user trials enabled to formulate clear recommendations on the role of the quality assessment in each phase of the SDL system design cycle. The SMEs tested the effectiveness of both the methodology of the tool and of the tool itself. This was done in the prospect of assessing and improving the SDL implementation process. They indicated the level of engineer acceptance of its findings. Quality assessment results presented for draft and validated SDL systems. Emphasis was given on whether the tool provides confirmation of good style practices. In addition, the added value of the automated quality assessment approach was compared to other methods for SDL systems quality improvement. The feedback and results from the tool evaluation task led to an optimised/error-free version of the SDL-SAFE tool. In this step, the qualitative performance of the tool was enhanced and its reliability was improved. Based on initial experience in the field trial, the feedback on the functionality and technical performance of the tool was exploited and the tool was optimised according to these results. Thus, the system was enhanced as necessary.
The SDL-SAFE project originates from the pressing requirement to test, measure and validate the quality and conformance of SDL generated code for critical communications and control systems. The project has reviewed this requirement in the context of the capabilities of Small and Medium sized Enterprises and seeked to provide a reliable and cost-effective solution. The objective of the Methodology document is to describe the methodology that leads to the automatic quality assessment of a software project implemented in SDL. A good practice is to follow a straightforward step-by-step approach. The steps of the methodology that will be followed are described here under. Initially, the style objectives for the proper use of SDL in SDL System implementations must be recommended. Thus, the style objectives will determine the way that an implementation should be developed in general. The second step is the definition of specific SDL language rules. This step is referenced to the style objectives, but it gives a more specific and strict image of how an SDL implementation should be developed in order to achieve high level of quality. The next step is the establishment of a quality assessment model, which will allow the quality assessment of an SDL implementation. J. A. McCall, P. K. Richards and G. F. Walters have developed a model especially for computer languages and notations. This model can also be applied on the quality assessment of an SDL system. The model consists of a number of factors and criteria. The fourth step to the quality assessment process is the extraction of metrics (measurement obtained without executing the SDL system). An automatic quality assessment process can be defined only based on numbers. The final step is the definition and explanation of the automatic quality assessment process. The process is based on the extracted metrics.

Searching for OpenAIRE data...

There was an error trying to search data from OpenAIRE

No results available