A BDD-based approach to verifying clone-enabled feature models' constraints and customization

Abstract

In this paper, we present a kind of semantics for constraints in clone-enabled feature models, which resolves the problem of what kinds of constraint should be added to a feature model after some features are cloned. The semantics is composed of two patterns: the generating pattern and the adapting pattern, to address the two problems of what kind of constraints should be imposed on a clonable feature and its clones, and how an existing constraint should be transformed in the context that features involved in the constraint are cloned, respectively. After that, we propose a BDD-based approach to verifying clone-enabled feature models, an approach that makes efficient use of the BDD (binary decision diagram) data structures, by considering the specific characteristics of feature models' verification. Experiments show that this BDD-based approach is more efficient and can verify more complex feature models than our previous method. © 2008 Springer-Verlag Berlin Heidelberg.