Software-intensive Systems Research Articles

MOSAICO: offline synthesis of adaptation strategy repertoires with flexible trade-offs

Self-adaptation improves the resilience of software-intensive systems, enabling them to adapt their structure and behavior to run-time changes (e.g., in workload and resource availability). Many of these approaches reason about the best way of adapting by synthesizing adaptation plans online via planning or model checking tools. This method enables the exploration of a rich solution space, but optimal solutions and other guarantees (e.g., constraint satisfaction) are computationally costly, resulting in long planning times during which changes may invalidate plans. An alternative to online planning involves selecting at run time the adaptation best suited to the current system and environment conditions from among a predefined repertoire of adaptation strategies that capture repair and optimization tasks. This method does not incur run-time overhead but requires additional effort from engineers, who have to specify strategies and lack support to systematically assess their quality. In this article, we present MOSAICO, an approach for offline synthesis of adaptation strategy repertoires that makes a novel use of discrete abstractions of the state space to flexibly adapt extra-functional behavior in a scalable manner. The approach supports making trade-offs: (i) among multiple extra-functional concerns, and (ii) between computation time and adaptation quality (varying abstraction resolution). Our results show a remarkable improvement on system qualities in contrast to manually-specified repertoires. More interestingly, moderate increments in abstraction resolution can lead to pronounced quality improvements, whereas high resolutions yield only negligible improvement over medium resolutions.

On the proposal and evaluation of a benchmark-based threshold derivation method

Software-intensive systems have been growing in both size and complexity. Consequently, developers need better support for measuring and controlling the software quality. In this context, software metrics aim at quantifying different software quality aspects. However, the effectiveness of measurement depends on the definition of reliable metric thresholds, i.e., numbers that characterize a metric value as critical given a quality aspect. In fact, without proper metric thresholds, it might be difficult for developers to indicate problematic software components for correction, for instance. Based on a literature review, we have found several existing methods for deriving metric thresholds and observed their evolution. Such evolution motivated us to propose a new method that incorporates the best of the existing methods. In this paper, we propose a novel benchmark-based method for deriving metric thresholds. We assess our method, called Vale’s method, using a set of metric thresholds derived with the support of our method, aimed at composing detection strategies for two well-known code smells, namely god class and lazy class. For this purpose, we analyze three benchmarks composed of multiple software product lines. In addition, we demonstrate our method in practice by applying it to a benchmark composed of 103 Java open-source software systems. In the evaluation, we compare Vale’s method to two state-of-the-practice threshold derivation methods selected as a baseline, which are Lanza’s method and Alves’ method. Our results suggest that the proposed method provides more realistic and reliable thresholds, with better recall and precision in the code smell detection, when compared to both baseline methods.

Supporting semi-automatic co-evolution of architecture and fault tree models

During the whole life-cycle of software-intensive systems in safety-critical domains, system models must consistently co-evolve with quality evaluation models like fault trees. However, performing these co-evolution steps is a cumbersome and often manual task. To understand this problem in detail, we have analyzed the evolution and mined common changes of architecture and fault tree models for a set of evolution scenarios of a part of a factory automation system called Pick and Place Unit. On the other hand, we designed a set of intra- and inter-model transformation rules which fully cover the evolution scenarios of the case study and which offer the potential to semi-automate the co-evolution process. In particular, we validated these rules with respect to completeness and evaluated them by a comparison to typical visual editor operations. Our results show a significant reduction of the amount of required user interactions in order to realize the co-evolution.

Special issue on “software quality in software-intensive systems”

Special issue on “software quality in software-intensive systems”

Substantially Evolutionary Theorizing in Designing Software-Intensive Systems

Useful inheritances from scientific experience open perspective ways for increasing the degree of success in designing of systems with software. One such way is a search and build applied theory that takes into account the nature of design and the specificity of software engineering. This paper presents a substantially evolutionary approach to creating the project theories, the application of which leads to positive effects that are traditionally expected from theorizing. Any implementation of the approach is based on a reflection by designers of an operational space of designing onto a semantic memory of a question-answer type. One of the results of such reflection is a system of question-answer nets, the nodes of which register facts of interactions of designers with accessible experience. A set of such facts is used by designers for creating and using the theory that belongs to the new subclass of Grounded Theories. This sub-class is oriented on organizationally behavioral features of a project’s work based on design thinking, automated mental imagination, and thought experimenting that facilitate increasing the degree of controlled intellectualization in the design process and, correspondingly, increasing the degree of success in the development of software-intensive systems.

The GRADE taxonomy for supporting decision-making of asset selection in software-intensive system development

ContextThe development of software-intensive systems includes many decisions involving various stakeholders with often conflicting interests and viewpoints. ObjectiveDecisions are rarely systematically documented and sporadically explored. This limits the opportunity for learning and improving on important decisions made in the development of software-intensive systems. MethodIn this work, we enable support for the systematic documentation of decisions, improve their traceability and contribute to potentially improved decision-making in strategic, tactical and operational contexts. ResultsWe constructed a taxonomy for documentation supporting decision-making, called GRADE. GRADE was developed in a research project that required composition of a common dedicated language to make feasible the identification of new opportunities for better decision support and evaluation of multiple decision alternatives. The use of the taxonomy has been validated through thirty three decision cases from industry. ConclusionThis paper occupies this important yet greatly unexplored research gap by developing the GRADE taxonomy that serves as a common vocabulary to describe and classify decision-making with respect to architectural assets.

Measuring object-oriented design principles: The results of focus group-based research

Object-oriented design principles are fundamental concepts that carry important design knowledge and foster the development of software-intensive systems with a focus on good design quality. They emerged after the first steps in the field of object-oriented programming and the recognition of best practices in using this programming paradigm to build maintainable software. Although design principles are known by software developers, it is difficult to apply them in practice without concrete rules to follow. We recognized this gap and systematically derived design best practices for a number of design principles and provide tool support for automatic measurement of these practices. The aim of this paper is to examine the relationship between design best practices and 10 selected design principles. This should provide evidence whether the key design aspects of the design principles are covered. We conducted focus group research with six focus groups and 31 participants in total. In parallel, each group discussed five design principles and assessed the coverage by using the Delphi method. Despite suggestions of additional design practices that were added by the participants, the result reveals the impact of each design best practice to the design principle and shows that the main design aspects of the design principles are covered by our approach and is therefore feasible to derive concrete design improvement actions.

Choosing Component Origins for Software Intensive Systems: In-House, COTS, OSS or Outsourcing?—A Case Survey

The choice of which software component to use influences the success of a software system. Only a few empirical studies investigate how the choice of components is conducted in industrial practice. This is important to understand to tailor research solutions to the needs of the industry. Existing studies focus on the choice for off-the-shelf (OTS) components. It is, however, also important to understand the implications of the choice of alternative component sourcing options (CSOs), such as outsourcing versus the use of OTS. Previous research has shown that the choice has major implications on the development process as well as on the ability to evolve the system. The objective of this study is to explore how decision making took place in industry to choose among CSOs. Overall, 22 industrial cases have been studied through a case survey. The results show that the solutions specifically for CSO decisions are deterministic and based on optimization approaches. The non-deterministic solutions proposed for architectural group decision making appear to suit the CSO decision making in industry better. Interestingly, the final decision was perceived negatively in nine cases and positively in seven cases, while in the remaining cases it was perceived as neither positive nor negative.

A framework for semi-automated co-evolution of security knowledge and system models

Security is an important and challenging quality aspect of software-intensive systems, becoming even more demanding regarding long-living systems. Novel attacks and changing laws lead to security issues that did not necessarily rise from a flawed initial design, but also when the system fails to keep up with a changing environment. Thus, security requires maintenance throughout the operation phase. Ongoing adaptations in response to changed security knowledge are inevitable. A necessary prerequisite for such adaptations is a good understanding of the security-relevant parts of the system and the security knowledge.We present a model-based framework for supporting the maintenance of security during the long-term evolution of a software system. It uses ontologies to manage the system-specific and the security knowledge. With model queries, graph transformation and differencing techniques, knowledge changes are analyzed and the system model is adapted. We introduce the novel concept of Security Maintenance Rules to couple the evolution of security knowledge with co-evolutions of the system model.As evaluation, community knowledge about vulnerabilities is used (Common Weakness Enumeration database). We show the applicability of the framework to the iTrust system from the medical care domain and hence show the benefits of supporting co-evolution for maintaining secure systems.

Variability and Complexity in Software Design

Today's software systems must accommodate a wide range of usage and deployment scenarios. The increasing size and heterogeneity of software-intensive systems, dynamic and critical operating conditions, fast moving and highly competitive markets, and increasingly powerful and versatile hardware makes it more and more difficult to handle the additional complexity in design caused by variability. This paper reports results of the Second International Workshop on Variability and Complexity in Software Design. It also outlines directions the field might move in the future.

Proactive model based testing and evaluation for component-based systems

Embedded software systems are getting more and more complex. The demand for new features and functions led to an increasing complexity in the design and development of these systems. There are frequent reports in the media about software systems crashing and damages occurring due to software errors. One reason for this is that there are many software testing methods and techniques but they are often non-practical and difficult to use. The aim of the study was to improve existing testing methods and their practicality especially from the integrator viewpoint. Component-based system development, components of different granularities must be tested. Furthermore, an optimization approach based on simulated annealing is presented which is used to derive an integration order with respect to the proposed parameters in a powerful and reliable manner. The paper discusses explicit properties and the requirements that are to be verified, imposed upon software-intensive systems by their environment and by their users.

Characterizing the contribution of quality requirements to software sustainability

BackgroundSince sustainability became a challenge in software engineering, researchers mainly from requirements engineering and software architecture communities have contributed to defining the basis of the notion of sustainability-aware software. ProblemDespite these valuable efforts, the assessment and design based on the notion of sustainability as a software quality is still poorly understood. There is no consensus on which sustainability requirements should be considered. Aim and MethodTo fill this gap, a survey was designed with a double objective: i) determine to which extent quality requirements contribute to the sustainability of software-intensive systems; and ii) identify direct dependencies among the sustainability dimensions. The survey involved different target audiences (e.g. software architects, ICT practitioners with expertise in Sustainability). We evaluated the perceived importance/relevance of each sustainability dimension, and the perceived usefulness of exploiting a sustainability model in different software engineering activities. ResultsMost respondents considered modifiability as relevant for addressing both technical and environmental sustainability. Functional correctness, availability, modifiability, interoperability and recoverability favor positively the endurability of software systems. This study has also identified security, satisfaction, and freedom from risk as very good contributors to social sustainability. Satisfaction was also considered by the respondents as a good contributor to economic sustainability.

Comparison of the FMEA and STPA safety analysis methods\u2013a case study

As our society becomes more and more dependent on IT systems, failures of these systems can harm more and more people and organizations. Diligently performing risk and hazard analysis helps to minimize the potential harm of IT system failures on the society and increases the probability of their undisturbed operation. Risk and hazard analysis is an important activity for the development and operation of critical software intensive systems, but the increased complexity and size puts additional requirements on the effectiveness of risk and hazard analysis methods. This paper presents a qualitative comparison of two hazard analysis methods, failure mode and effect analysis (FMEA) and system theoretic process analysis (STPA), using case study research methodology. Both methods have been applied on the same forward collision avoidance system to compare the effectiveness of the methods and to investigate what are the main differences between them. Furthermore, this study also evaluates the analysis process of both methods by using a qualitative criteria derived from the technology acceptance model (TAM). The results of the FMEA analysis were compared to the results of the STPA analysis, which were presented in a previous study. Both analyses were conducted on the same forward collision avoidance system. The comparison shows that FMEA and STPA deliver similar analysis results.

Future software organizations \u2013 agile goals and roles

Digital transformation is rapidly causing major, even disruptive changes in many industries. Moreover, global developments like digital platforms (cloud) and IoT create fundamentally new connections at many levels between objects, organizations and people (systems-of-systems). These are by nature dynamic and often work in real time – further increasing the complexity. These systemic changes bring up new profound questions: What are those new software-intensive systems like? How are they created and developed? Which principles should guide such organizational design? Agile enterprises are by definition proficient with such capabilities. What solutions are the current scaled agile frameworks such as SAFe and LeSS proposing, and why? In this paper, we aim to recognize the design principles of future software organizations, and discuss existing experiences from various different organizations under transformations, and the insights gained. The purpose is to systematize this by proposing a competence development impact-mapping grid for new digitalization drivers and goals with potential solutions based on our agile software enterprise transformation experiences. Our research approach is based on the resource-based and competence-based views (RBV, CBV) of organizations. We point out how most decision-making in companies will be more and more software-related when companies focus on software. This has profound impacts on organizational designs, roles and competencies. Moreover, increasing data-intensification poses new demands for more efficient organizational data processing and effective knowledge utilization capabilities. However, decisive systematic transformations of companies bring new powerful tools for steering successfully under such new business conditions. We demonstrate this via real-life examples.

A Practical Analytical Approach to Increase Confidence in Software Safety Arguments

Complex interdependencies among a large and growing number of elements in computer-based critical systems have made the safety analysis process more burdensome in terms of costs and tend to limit the capabilities of current safety analysis approaches to address system hazards. Therefore, the critical systems lifecycle—primarily software-intensive systems—demands new approaches for conducting the safety analysis process while staying within reasonable cost limits and maintaining the quality of the results. This paper proposes a practical analytical approach for increasing the confidence in software safety arguments in safety cases through the elaboration and assessment of counterevidence that emerges from software failure modes due to robustness issues. A case study is presented in which the proposed approach is applied to a computer-based distributed critical system. Additionally, practical results from the application of this approach to real safety-critical systems are also presented. The proposed approach proved to be useful during safety analysis processes, mainly in independent safety analysis projects, as it improves assessment coverage with low impact on workload and costs.

A tour of secure software engineering solutions for connected vehicles

The growing number of vehicles daily moving on roads increases the need of protecting the safety and security of passengers, pedestrians, and vehicles themselves. This need is intensified when considering the pervasive introduction of Information and Communication Technologies (ICT) systems into modern vehicles, because this makes such vehicles potentially vulnerable from the point of view of security. The convergence of safety and security requirements is one of the main outstanding research challenges in software-intensive systems. This work reviews existing methodologies and solutions addressing security issues in the automotive domain with a focus on the integration between safety and security aspects. In particular, we identify the main security issues with vehicular communication technologies and existing gaps between state-of-the-art methodologies and their implementation in the real world. Starting from a literature survey and referring to widely accepted standards of the domain, such as AUTOSAR and ISO 26262, we discuss research challenges and set baselines for a holistic secure-by-design approach targeting safety and security aspects all along the different phases of the development process of automotive software.

A multi-scale modeling approach for systems of systems architectures

Modeling correct software-intensive Systems of Systems architectures is a challenging research direction that can be mastered by providing appropriate modeling abstractions. For this purpose, we provide an iterative modeling solution for a multi-scale description of software architectures. We provide a visual notation extending the graphical UML notations to represent structural as well as behavioral features of software architectures. We define a step-wise iterative process from a coarse-grain to a fine-grain description. The intermediate iterations provide a description with a given abstraction that allow the validation to be conducted significantly while remaining tractable w.r.t. complexity. The iterative process involves both system-independent structural features ensuring the model correctness, and system-specific features related to the expected behavior of the modeled domain. We implement the resulting models describing structural and behavioral properties using the Eclipse platform. To experiment our approach, we deal with an SoS dedicated to the Emergency Response and Crisis Management System(ERCMS).

Application of SPN Simulation in Reliability Evaluation of Software-intensive Systems

This paper presents a model based on Stochastic Petri Net, which evaluates the reliability of software-intensive systems, and analyzes the relationship between the hardware and software, and establishes the simulation model according to the fault performance. This model solves the exponential explosion problem of stochastic Petri net when it is converted into continuous time Markov chain. Several performance of software-intensive reliability is estimated.

Stimuli-SoS: a model-based approach to derive stimuli generators for simulations of systems-of-systems software architectures

BackgroundSystems-of-systems (SoS) are alliances of independent and interoperable software-intensive systems. SoS often support critical domains, being required to exhibit a reliable operation, specially because people’s safety relies on their services. In this direction, simulations enable the validation of different operational scenarios in a controlled environment, allowing a benchmarking of its response as well as revealing possible breaches that could lead to failures. However, simulations are traditionally manual, demanding a high level of human intervention, being costly and error-prone. A stimuli generator could aid in by continuously providing data to trigger a SoS simulation and maintaining its operation.MethodsWe established a model-based approach termed Stimuli-SoS to support the creation of stimuli generators to be used in SoS simulations. Stimuli-SoS uses software architecture descriptions for automating the creation of such generators. Specifically, this approach transforms SoSADL, a formal architectural description language for SoS, into dynamic models expressed in DEVS, a simulation formalism. We carried out a case study in which Stimuli-SoS was used to automatically produce stimuli generators for a simulation of a flood monitoring SoS.ResultsWe run simulations of a SoS architectural configuration with 69 constituent systems, i.e., 42 sensors, 9 crowdsourcing systems, and 18 drones. Stimuli generators were automatically generated for each type of constituent. These stimuli generators were capable of receiving the input data from the database and generating the expected stimuli for the constituents, allowing to simulate constituent systems interoperations into the flood monitoring SoS. Using Stimuli-SoS, we simulated 38 days of flood monitoring in little more than 6 h. Stimuli generators correctly forwarded data to the simulation, which was able to reproduce 29 flood alerts triggered by the SoS during a flooding event. In particular, Stimuli-SoS is almost 65 times more productive than a manual approach to producing data for the same type of simulation.ConclusionsOur approach succeeded in automatically deriving a functional stimuli generator that can reproduce environmental conditions for simulating a SoS. In particular, we presented new contributions regarding productivity and automation for the use of a model-based approach in SoS engineering.

Characterization of trade-off preferences between non-functional properties

Efficient design and evolution of complex software intensive systems rely on the ability to make informed decisions as early as possible in the life cycle. Such informed decisions should take both the intended functional and non-functional properties into account. Especially regarding the latter, it is both necessary to be able to predict properties and to prioritize them according to well-defined criteria. In this paper we focus on the latter problem, that is to say how to make trade-offs between non-functional properties of software intensive systems. We provide an approach based on the elicitation of utility functions from stake-holders and subsequent checks for consistency among these functions. The approach is exploitable through an easy-to-use GUI, which is also presented. Moreover, we describe the setup and the outcome of our two-fold validation based on exploratory elicitations with students and practitioners.