Critivar: a variability modeling and transformation approach for safety and security-critical systems

Abstract

Mass customization of modern critical systems has led to products with many variation points, impacting their architecture and dependability properties. Given the rudimentary vulnerability management techniques available, modeling diversity in critical systems and their corresponding dependability (e.g., safety and security) characteristics is challenging. That being said, this dissertation presents a novel annotative variability management approach for critical systems. The approach is illustrated using a case study from the automotive domain. As part of the evaluation, the proposed approach is compared against two similar vulnerability management solutions. Among the benefits, the the proposed approach reduces the gap between variability and finer-grained model dependability annotations and ensures the derivation of correct and complete critical system models.