Handling multiple quality attributes trade-off in architecture-based self-adaptive systems

Self-adaptive systems are capable of autonomously making runtime decisions in order to deal with uncertain circumstances. In architecture-based self-adaptive (ABSA) systems the feedback loop uses self-reflecting models to perform decision making and ultimately apply adaptation to the system. One aspect of this decision making mechanism is to handle systems' quality attributes trade-off. An ABSA system is required to address the potential impacts of adaptation on multiple quality attributes, and select the adaptation option which satisfies the quality attributes of the system the best. In this PhD project, we study and propose an architecture-based solution which uses runtime knowledge of the systems and its environment to handle quality attributes trade-off and decision making mechanism in presence of system's quality goals uncertainty. For validation, we will a) create and set up case studies in various domains, and b) use exemplars to benchmark our proposed method with existing approaches.