Model-Based Approach for Co-optimization of Safety and Security Objectives in Design of Critical Architectures

Abstract

During the development of Cyber-Physical Systems (CPS) safety and security are major concerns to be considered as it has been established by various literature. Moreover, these concerns must be included early on during the System Development Life Cycle (SDLC). In this work, we focus on the design-phase of the SDLC to assist the engineers in conducting design-space exploration of the system hardware architecture w.r.t to both safety and security concerns. In this way, the engineers may perform simulations to find a set of quasi-optimal solutions before developing an actual physical prototype. To achieve this, our tooled method builds on our previous work [11] and supports a multiconcern analysis by leveraging Model-Driven Engineering (MDE) techniques such as SysML modeling along with the transformation of SysML models into representations which are finally optimized via constraint solvers. Overall, the method and framework shall support the design of the system architecture from a repository of components based on possible configuration alternatives, which satisfy the system objectives such as reliability and cost. Such functions can help to evaluate the effects of integrating safety and security features thus showing their interplay. The overall approach is illustrated via an automotive CPS case study.