Designing and developing complex self-adaptive systems require design processes having specific features fitting and representing the complexity of these systems. Changing requirements, users’ needs and dynamic environment have to be taken in consideration, also considering that, due of the self-adaptive nature of the system, the solution is not fixed at design time but it is a run-time outcome. Traditional design approach and life cycles are not suitable to design software systems where requirements continuously change at runtime. A new design process paradigm is needed to design such systems. In this Chapter, we present a retrospective analysis based on three projects developed in the last five years with the middleware MUSA in order to identify specific features of the design process for supporting continuous change and self-adaptation. The result is a general approach allowing to reduce the gap between design time and run-time.
|Titolo della pubblicazione ospite||Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)|
|Numero di pagine||21|
|Stato di pubblicazione||Published - 2019|
|Nome||LECTURE NOTES IN ARTIFICIAL INTELLIGENCE|