- Systems Modeling Language (SysML) is used for specifying, analyzing, designing and verifying complex systems, and is designed to provide simple but powerful constructs for modeling a wide range of Systems. SysML is not a methodology, nor a method and there are thousands of different ways to describe the using it. In this case, there cannot be a single, universal approach to evaluate the consistency of the requirements specification. It is necessary to choose a specific method in combination with SysML to accurately and comprehensively evaluate the consistency of requirements specification. The consistency evaluation of requirements specification in modelbased system engineering (MBSE), depending on the modeling language and method is quite a new practice. This opens up discussions of how to utilize SysML provided infrastructure to evaluate the System Requirements Specification (SRS) and achieve a high-quality of the SRS. In this paper, a new approach of how requirements specification, expressed with sufficient precision in SysML can be used for automated consistency evaluation.
Keywords—SysML, MBSE Grid, Consistency Metrics, System Requirements Specification, Requirements Engineering, MBSE
Due to model-based engineering progress in recent years, system engineering slowly but surely moves from documentbased system engineering to model-based system engineering. Nowadays, MBSE is enabled by Systems Modeling Language.
SysML is a general-purpose graphical modeling language
that supports the analysis, specification, design, verification, and
validation of complex systems. The language is intended to
create cohesive and consistent models of structure, behavior
including their interconnections. SysML introduces requirement
diagrams, which represent requirements and their relationships
to other requirements, design elements and test cases [
Requirements engineering is one of the most important and critical phases in MBSE which consists of two main processes: specification and management. Generally, for systems engineers are more important requirements specification process than requirements management process. While managers focus is on the requirements management process, but they have a poor understanding of the benefits of MBSE. According to PMI's
study, poor requirements management is the second most
common reason for project failure [
In order to reduce the risk of mistakes detection and
correction in the late stage of development, it is desirable and
important to identify the inconsistencies in a requirements
specification in the early stages of systems engineering. The
mistakes due to incompleteness, inconsistency, and ambiguity
introduced at the stage of requirements engineering are difficult
and more expensive to correct than those introduced in later
stages of system development [
In this paper, we focus on a subset of the C&C task – correctness analysis only. We understand the correctness of SRS as the nonexistence of inappropriate relationships between requirements and model elements. The question is how to utilize SysML provided infrastructure to successfully achieve a high quality of the requirements specification: what method to use in combination with SysML.
In this paper, we propose a new approach of how requirements specification that is expressed in SysML in combination with MBSE Grid method can be used for automated consistency evaluation of the system requirements specification.
The MBSE Grid method guides how to specify principal
areas of the system model and how to manage different layers of
abstraction [
The MBSE Grid is organized in a matrix view. Columns
represent four main aspects of systems engineering
(requirements, system structure, system behavior and
parameters). Rows represent two main viewpoints: one to define
the problem in order to understand it, other to provide one or
several alternative solutions to solve it. Cells of the grid (Fig. 1)
represent different views of model-based systems engineering
[
This research is carried out using MagicDraw toolset, which
supports SysML. It was chosen because of several published
studies, e.g. [
The rest of this paper is structured as follows: in section 2, the related works are analyzed; in section 3, the proposed approach for automated consistency evaluation of the requirements specification is presented; in section 4, evaluation of the proposed approach is described; in section 5, the achieved results, conclusions, and future work directions are indicated.
There is a large number of research papers on the consistency
analysis of requirements specification. Most of them are applied
to the small area of the domain or a specific tool, e.g. [
Several authors proposed methodologies for evaluation of
consistency within the UML models which are applicable to
SysML, as SysML is the extension of UML. Methods defined in
[
Use of traceability relationships to evaluate the consistency
of the requirements specification has been defined in [
In [
In conclusion, all the analyzed methods to evaluate the consistency of requirements specification encounter several common issues: (i) unclear traceability relationships between requirements and design elements, (ii) unsupported consistency evaluation at all stages of the requirements specification, (iii) it is difficult to interpret the results of consistency evaluation.
Overall, researches carried out in this area have very little proof of its successful application on real-industry cases or is very specific to a small area of application and specific tools dependent. We are proposing a more generic, easy to use approach, applicable to the majority of SysML modeling tools for different systems engineering domains. The proposed approach in combination with MBSE Grid will evaluate the consistency of each stage of requirements specification. This will help to monitor the quality of SRS and make the necessary decisions in the early stage of requirements specification process.
III. AN APPROACH FOR CONSISTENCY EVALUATION OF SYSTEM
REQUIREMENTS SPECIFICATIONS This section describes the proposed approach in detail.
The approach consists of the following metric groups that are defined on the basis of the principles of MBSE Grid method:
The metric groups mentioned above compute only atomic model elements that are linked to the atomic requirements (child requirement). The relation between atomic requirements and atomic model elements eliminates the ambiguities that may occur having relations between higher level elements.
Fig. 2. MBSE Grid Traceability
The approach is based on specified traceability relationships in MBSE Grid [Fig. 2]. Requirements Refinement Metrics compute the consistency of requirements specification at problem layer. This metrics group evaluates the refinement of stakeholder needs by elements that are specified at stages of functional analysis, logical subsystems communication and measurements of effectiveness of Subsystems (MoES). Requirements Satisfaction Metrics compute the consistency of requirements specification at solution layer. This metrics group evaluates system requirements satisfaction by elements that are specified at stages of component behavior, structure and parameters. Requirements Verification Metric evaluates the consistency between system requirements and test cases.
The proposed method concerns the consistency evaluation of the system requirements specifications. An approach is implemented in the MagicDraw modeling tool.
In order to obtain the more precise evaluation results of requirements specification, metrics are categorized by three aspects of system engineering: Behavior, Structure, and Parameters.
This metric group evaluates the consistent use of model elements to refine stakeholder needs. The stakeholder needs are refined by the behavior elements specified in the functional analysis, by the structure elements specified in the logical subsystem communication and by the parameters specified at the measurements of effectiveness.
The metric group of requirements refinement consists of the following metrics:
This metric evaluates the utilization of behavior elements to refine requirements. The atomic activity elements defined in the functional analysis have to refine the atomic functional requirements of stakeholder. Below is provided the metric formula. This metric evaluates the utilization of structure elements to refine requirements. The atomic block or part elements defined in the logical subsystem communication have to refine the atomic physical requirements of stakeholder. Below is provided the metric formula.
ℎ
= This metric evaluates the utilization of proxy elements to refine requirements. Proxy ports defined in the logical subsystem communication have to refine the atomic interface requirements of stakeholder. Below is provided the metric formula.
= This metric evaluates the utilization of behavior elements to satisfy requirements. The atomic activity elements defined in the component behavior have to satisfy the atomic functional requirements of the system. Below is provided the metric formula.
=
MESBE - functional requirements satisfaction by behavior elements metric BESFR – quantity of behavior elements used to satisfy functional
requirements behavior analysis BE – quantity of behavior elements defined in the component defined in the component assembly have to satisfy the atomic physical requirements of the system. Below is provided the metric formula.
ℎ
= This metric evaluates the utilization of proxy elements to satisfy requirements. Proxy ports defined in the component assembly have to satisfy the atomic interface requirements of the system. Below is provided the metric formula. = PPESIR – quantity of proxy port elements used to satisfy functional PPE – quantity of proxy port elements defined in the component The metric evaluates the utilization of value property elements to satisfy requirements. The value property elements defined in the component parameters analysis have to satisfy the atomic performance requirements of the system. Below is provided the metric formula.
= verify the system requirements. Defined test cases have to verify atomic system requirements. Below is provided the metric formula.
= refinement evaluation of stakeholder needs applying the requirements refinement by behavior elements metric (1).
The figure below (Fig. 3) represents the stakeholder needs refinement by atomic activity element.
First, we calculate the quantity of atomic activity elements that are
defined in the functional analysis.
Activities have to represent the behavior of the system.
Second, we calculate the quantity of atomic activity elements that are used to refine the atomic functional requirements of stakeholder.
Third, we calculate the evaluation of requirement refinement by behavior elements using the particular metric.
Below is provided the result of the evaluation of stakeholder need refinement according to Fig. 3.
BERFR = 1 BE = 2 1 2 = × 100% = 50%
This indicates that 50% of the activities which are specified at the stage of function analysis are used to refine the stakeholder needs. started the refinement of performance requirements stage. When all metrics reached over 90%, it was decided that the refinement of requirements specification is sufficient.
This section describes the case study of the proposed approach. This is a case study of a commercial project to evaluate the consistency of the requirements specification.
The following is a detailed description of the consistency analysis of requirements specification. The commercial project is based on SysML and is modeled in the MagicDraw toolset. The modeling carried out in accordance with the principles of MBSE grid.
Requirements specifications Consistency metrics have been computed over the entire period of SRS. After each metric calculation, the responsible persons have been analyzed the metrics data and made appropriate decisions to ensure a high quality of the SRS.
Fig. 4 shows the part of requirements satisfaction metric table that is computed in the MagicDraw tool. For effective analysis, metrics data was exported to the excel and the visual charts were created according to the metrics data.
Below is provided a detailed analysis of each metric groups that are presented in the charts.
In Fig. 5 is displayed refinement analysis diagram of requirements specification. Requirements refinement metrics have been computed over a period specifying the problem layer of requirements specification. First, the functional requirements of stakeholder have been refined by behavior elements. Reaching the 85% of behavior elements usage for refining functional requirements of stakeholder has been started another stage, the refinement of physical and interface requirements. Reaching over the 80% of proxy ports usage for refining interface requirements of stakeholder and structure elements usage for refining physical requirements of stakeholder has been
Fig. 6. Requirements Satisfaction Diagram
In Fig. 6 is displayed satisfaction analysis diagram of requirements specifications over a period specifying the solution layer of requirements specification. First, the functional requirements of the system have been satisfied by behavior elements. Reaching the 82% of behavior elements usage for satisfying the functional requirements of the system has been started another stage, the satisfaction of physical and interface requirements. Reaching over the 85% of proxy ports usage for satisfying the interface requirements and structure elements usage for satisfying the physical requirements has been started other stage, the satisfaction of performance requirements. When all metrics reached over 90%, it was decided that the satisfaction of requirements specification is sufficient.
The analysis of existing consistency evaluations methods for the requirements specification disclosed that there are multiple different methods available. The majority of them cannot be used in combination with systems modeling techniques, such as SysML, in practice. We found a need to propose a more generic, easy to use approach, applicable to the majority of SysML modeling tools for different system engineering domains.
In this paper, we proposed a new approach of how requirements specification, expressed with sufficient precision in SysML, can be used for automated consistency evaluation. The approach consists three metric groups that are defined on the basis of the principles of MBSE Grid method: Requirements Refinement Metrics, Requirements Satisfaction Metrics, Requirements Verification Metric.
We have implemented the proposed approach in the MagicDraw CASE tool and demonstrated an example case study. After analyzing the case study, it was determined that calculation of the consistency metrics over a period contributes to ensure a high quality of each stage of requirements specification.
Currently, the approach is oriented to automated consistency evaluation of requirements specification. However, we plan to extend the approach in the near future, to evaluate the completeness of requirements specification and, finally, we seek to combine both to evaluate more precisely the requirements specification in model-based systems engineering.