Model-driven synthesis of formally precise, stylized software architectures

Abstract

Reliably producing software architectures in selected architectural styles requires significant expertise yet remains difficult and error-prone. Our research goals are to better understand the nature of style-specific architectures, and relieve architects of the need to produce such architectures by hand. To achieve our goals, this paper introduces a formally precise approach to separate architectural style design decisions from application-specific decisions, and then uses these separate decisions as inputs to an automated synthesizer. This in effect supports a model-driven development (MDD) approach to architecture synthesis with style as a separate design variable. We claim that it is possible to formalize this separation of concerns, long implicit in software engineering research; to automatically synthesize style-specific architectures; and thereby to improve software design productivity and quality. To test these claims, we employed a combination of experimental systems and case study methods: we developed an MDD tool and used it to carry out case studies using Kitchenham’s methods. Our contributions include: a theoretical framework formalizing our separation of concerns and synthesis approach; an MDD framework, Monarch; and results of case studies that we interpret as supporting our claims. This work advances our understanding of software architectural style as a formal refinement; makes application descriptions an explicit subject of study; and suggests that synthesis of architectures can improve software productivity and quality.

Why is My Component and Connector Views Specification Unsatisfiable?

“…Bagheri and Sullivan [3], [4] present a different approach to using Alloy for specifying structural properties of component and connector models. To the best of our understanding, they do not handle a case of unsatisfiability.…”

Platinum: Reusing Constraint Solutions in Bounded Analysis of Relational Logic

“…Software verification provides the highest degree of software assurance, with its strengths residing in the mathematical concepts that can be leveraged to prove correctness with respect to specific properties. Most notably, bounded verification techniques, such as Alloy [28], have recently received a great deal of attention in the software engineering community (e.g., [8,9,11,13,14,16,20,26,34,35,38,43,46,48,52,54,55,61,63,66]), due to the strength of their automated, yet formally precise, analysis capabilities. The basic idea behind these techniques is to construct a formula that encodes the behavior of a system and examine it up to a user-specified bound.…”