Applying Digital Evolution to the Development of Self-Adaptive ULS Systems

Abstract

A key characteristic for ultra-large scale (ULS) software- intensive systems is the need to adapt at run time in response to changing environmental conditions. Given the scale, complexity, and heterogeneity of ULS elements, innovative, but rigorous software engineering techniques are needed to address the development and the evolution of these systems. The developer of self-adaptive ULS systems must anticipate how and when the software will need to adapt in the future, codify this behavior in decision-making components to govern the adaptation, and ensure that system integrity is not compromised during adaptations. We contend that the full potential of dynamically adaptive software systems cannot be realized without environments that enable the developer to actively explore the "adaptation space " of the system during the early stages of design. We propose an approach to this problem that leverages and extends digital evolution techniques. By mapping models of adaptive software programs into digital organisms and studying traces of their evolution, the developer can gain critical insight into software decision making, software assurance, and the software infrastructure needed to support desired adaptations.