Selecting Architecture Configurations in Self- Adaptive Systems Using QoS Criteria

Abstract

The process for choosing the best architectural elements to build systems has become increasingly complex. It becomes more critical if it is necessary to consider many combinations of components in the context of an architectural configuration. This paper proposes an approach for the dynamic selection of architectural configurations based on quality of system criteria of the system, which can be used during a dynamic adaptation, the MoSAC. This module uses the MAUT theory (Multi-Attribute Utility Theory) for decision making from a finite set of possibilities, which involve multiple criteria to be analyzed. We also present a case study that is an application in the context of Iterative Digital TV to evaluate the time spent on the module to return a valid configuration to be used in a middleware with self-adaptive features, the middleware FRAME.