An Overview on Quality Evaluation of Self-Adaptive Systems

Evaluation is an important concern for building and monitoring the quality of software. The complex nature of self-adaptive systems demands continuous monitoring of their behavior and execution environment, which challenge the quality of their adaptations in dynamic environments. The characteristics of self-adaptive systems demand a continuous evaluation of their performances and improvement of the adaptation process. In this chapter our objective is to identify general guidelines for the evaluation of self-adaptive systems independent of their type, application domain, or implementation details. We propose a taxonomy for the evaluation of the quality of self-adaptive systems based on five dimensions: scope, time, mechanisms, perspective, and type. Further, we identify and introduce the main available evaluation approaches and analyze them based on the proposed taxonomy. Finally, we discuss several trade-offs concerning each dimension in the taxonomy, trade-offs which should be addressed during the evaluation of self-adaptive systems.