Stability of Self-Adaptive Software Architectures

Abstract

Stakeholders and organisations are increasingly looking for long-lived software. As architectures have a profound effect on the operational life-time of the software and the quality of the service provision, architectural stability could be considered a primary criterion towards achieving the long-livety of the software. Architectural stability is envisioned as the next step in quality attributes, combining many inter-related qualities. This research suggests the notion of behavioural stability as a primary criterion for evaluating whether the architecture maintains achieving the expected quality attributes, maintaining architecture robustness, and evaluating how well the architecture accommodates run-time evolutionary changes. The research investigates the notion of architecture stability at run-time in the context of self-adaptive software architectures. We expect to define, characterise and analyse this intuitive concept, as well as identify the consequent trade-offs to be dynamically managed and enhance the self-adaptation process for a long-lived software.