A Theoretical Framework for Self-Adaptive Systems: Specifications, Formalisation, and Architectural Implications
Ana Petrovska, Thomas Hutzelmann, Stefan Kugele · 2023
In the last two decades, the popularity of self-adaptive systems has significantly increased in the literature across various research domains. However, despite the extensive work in this research field and the vibrant research community, the literature still needs a more precise understanding of self-adaptive systems, including clear specifications and usable definitions. So far, the terminology related to self-adaptive systems is still used liberally, which can result in different entities having a different understanding of what these systems are. Having an ambiguous description of the terminology and the lack of a precise understanding of these systems adds complexity to an already complex field of research. As a result, the overarching objective of this paper is to broaden the understanding of self-adaptive systems. For improved clarity and coherence--- potentially yielding to even higher prominence of this domain---we consider as essential to set a common terminology and gain a better comprehension of the semantics of that terminology. In response, in this paper, we propose a theoretical framework for self-adaptive systems, in which we specify the process of self-adaptation, elicit the minimum requirements for self-adaptive systems, formalise these systems, and discuss various architectural implications from our theoretical contributions. Ideally, this work 1) sets a foundation for more constructive scientific debates, 2) complements the already existing works in this field, and 3) sets the foundation for future engineering approaches for self-adaptive systems.