Antonia Bertolino

Redundancy is widely used as a method for fault and intrusion tolerance. However, if the redundant components lack sufficient diversity, potentially dangerous common mode failures may go undetected. To address this issue, the design diversity approach has been proposed in the literature for decades. In this article, we take an innovative approach to this problem by introducing a broader notion of diversity, which leverages Metamorphic Relations (MRs), i.e., necessary properties that must hold among diverse inputs and diverse outputs. We define two generic categories of MRs that establish data diversity and functional diversity. Furthermore, we elaborate on two corresponding logical architectures, paying particular attention to the necessary conditions for the adjudicator component. Finally, we present an initial evaluation of the proposed architectures, which points out the advantages with respect to their counterparts based on the traditional design diversity method, and discuss future research directions for this novel conceptual approach to redundancy-based fault/intrusion tolerance.

In software testing, test suite augmentation refers to the activity of adding new test cases to an existing test suite to enhance its effectiveness, particularly in terms of code coverage and fault detection. In this work, we address the problem of test suite augmentation and propose to leverage implementation dependent test cases that are generated from the coverage of equivalent programs. Specifically, we define cross-coverage testing and cross-coverage test suite augmentation as well as their reference scenarios. Also, we empirically investigate cross-coverage test suite augmentation against two different subjects: a reference benchmark of functionally equivalent programs, and a real-world case study in the context of XACML-based access control systems. Our results show an improvement of the thoroughness of our cross-coverage augmented test suite, as well as of its fault-detection effectiveness, outperforming existing test augmentation strategies.

For more than a decade, software engineering for systems‐of‐systems (SoS) and software ecosystems (SECO) has been largely investigated in order to cope with complexity in software‐intensive systems. SoS research addresses several aspects related to software system architecture comprising a set of constituent systems that relate to each other to perform missions. As such, SoS have key characteristics such as operational and managerial independence, distribution, emergent behavior, and evolutionary development. Full interoperability and dynamic architecture become critical challenges in this context. On the hand, SECO research refers to modeling and analysis of a socio‐technical network of actors and artifacts formed on top of common technological platforms, in which business factors directly influence software maintenance and evolution. Software sustainability and diversity as well as quality attributes that affect the SECO platform health represent challenges in the field. From the long‐running, successful series of the International Workshop on Software Engineering for systems‐of‐systems and Software Ecosystems (SESoS), co‐located with the IEEE/ACM International Conference on Software Engineering (ICSE), we present this special issue on the topics in the Journal of Software: Evolution and Process from SESoS 2023 in Melbourne, Australia. Four articles were accepted and published in this special issue, covering a longitudinal analysis of SoS research, as well as strategic patterns, services, and trust in SECO. These articles provide researchers and practitioners with advances in the state of the art and point out opportunities for further research.

For more than a decade, software engineering for systems-of-systems (SoS) and software ecosystems (SECO) has been largely investigated in order to cope with complexity in software-intensive systems. SoS research addresses several aspects related to software system architecture comprising a set of constituent systems that relate to each other to perform missions. As such, SoS have key characteristics such as operational and managerial independence, distribution, emergent behavior, and evolutionary development. Full interoperability and dynamic architecture become critical challenges in this context. On the hand, SECO research refers to modeling and analysis of a socio-technical network of actors and artifacts formed on top of common technological platforms, in which business factors directly influence software maintenance and evolution. Software sustainability and diversity as well as quality attributes that affect the SECO platform health represent challenges in the field. From the long-running, successful series of the International Workshop on Software Engineering for Systems-of-Systems and Software Ecosystems (SESoS), co-located with the IEEE/ACM International Conference on Software Engineering (ICSE), we present this special issue on the topics in the Journal of Software: Evolution and Process from SESoS 2023 in Melbourne, Australia. Four articles were accepted and published in this special issue, covering a longitudinal analysis of SoS research, as well as strategic patterns, services, and trust in SECO. These articles provide researchers and practitioners with advances in the state of the art and point out opportunities for further research.