Software architecture is critical to the success of large software systems. It has long been recognized that architectural design has profound impact on the achievement of system requirements. Two typical methods have been proposed to help users to make architectural decisions based on requirements. One method uses risk-based reasoning to help choosing good architectural design that can meet both system requirements and resource limitations, the other one, using quality attribute model as the linkage, enables deriving a fragment of architectural design that is focused on achieving certain quality requirement. However, there is little effort on comparing the two methods to discover similarities and differences. In this paper, we conduct a systematic analysis of the two methods, and compare them from the aspects of system requirements, architectural decisions, and mapping approach. We then propose a procedure to combine the two methods that provides better support to architectural design decision making.
Date of Conference: 26-29 Oct. 2010