Dealing with Known Unknowns: Towards a Game-Theoretic Foundation for Software Requirement Evolution

Requirement evolution has drawn a lot of attention from the community with a major focus on management and consistency of requirements. Here, we tackle the fundamental, albeit less explored, alternative of modeling the future evolution of requirements. Our approach is based on the explicit representation of controllable evolutions vs observable evolutions, which can only be estimated with a certain probability. Since classical interpretations of probability do not suit well the characteristics of software design, we introduce a game-theoretic approach to give an explanation to the semantic behind probabilities. Based on this approach we also introduce quantitative metrics to support the choice among evolution-resilient solutions for the system-to-be. To illustrate and show the applicability of our work, we present and discuss examples taken from a concrete case study (the security of the SWIM system in Air Traffic Management). © 2011 Springer-Verlag.