Towards improving security testability of AADL architecture models

Today computer systems are becoming very complex and often depending on some off-the-shelf components or external service providers. Validating the security aspects of these systems is a highly challenging goal, especially when we target security critical systems where guarantees must be provided on the robustness of the produced systems. Model Driven Engineering (MDE) is becoming popular and well accepted in both academia and critical systems industry. In our work, we aim to propose a novel MDE compliant security validation methodology based on architecture model based testing. In this paper, we focus on security modeling for improving the testability of the architecture models and facilitating the evaluation of the proposed test generation and selection methods. More specifically, we are interested in extending the SAE standard AADL [1] to improve the quality of the security test cases generated from AADL models.