Abstract

In recent years, software ecosystems (SECO) have emerged as a key paradigm of modern software development, bringing together a large number of entities that collaborate to create, maintain and evolve software products. While contributing to collaboration, innovation, and quick software evolution, SECO also pose com-plex architectures that may impact on scalability, security, reliability, flexibility, interface stability, and long term sustainability. Although a previous Systematic Literature Review (SLR) identified key architectural challenges and how they might be mitigated in SECO, its relative importance and the order in which to consider its mitigation strategies was unclear. To fill this gap, the present study emphasizes on architectural challenges and the corresponding mitigation strategies for the events identified through analysis of SLR, using Multi-Criteria Decision-Making (MCDM) approach Analytic Network Process (ANP) which is appropriate for complex decision variables. The expert judgment data were gathered by carrying out a structured survey from software practitioners of the software ecosystem and professionals and then analyzed using the ANP based pairwise comparison and priority weight computation. The results indicate that the most important architectural challenge is Scalability of SECO Architecture, then Security, Architecture Flexibility, Reliability of SECO Architecture, and lastly Interface stability. Based upon the above, the capacity of software ecosystems to assist in growth, interaction assurance, adaptability to architecture, service reliability and stable interfaces are key to sustainable development of a software eco-system. With regard to mitigation strategies, Framework is the top performing mitigation methodology, followed by Metrics, Custom Design, Assessment of Ecosystem Health, Sandbox, and Variable Design Decisions. This study further adds value to the software ecosystem research field by adding the findings of the previously mentioned SLR which was published in the field, through quantitative prioritization. The proposed prioritization based on ANPs offers pragmatic suggestions for SECO software architects, researchers, and practitioners towards identifying the most important architectural challenges and choosing corrective actions, i.e., mitigation strategies. The results assist in making informed architectural decisions, proactive risk management, and enhancing the sustainability of software ecosystem architecture.

Authors

Inayat Ur Rahman¹, Sahib Bahadar², Sara Shahzad³
Nanjing University of Information Science and Technology, China^1,2, University of Peshawar, Pakistan³

Keywords

SECO, ANP, MCDM, NumPy, Software Ecosystem Architecture