- Verify, detect and monitor IoT components and provide self-test functionality to them to ensure correctness and robustness of the whole CPS with respect to their formalized security requirements
- Automatize generation of test cases based on model-mutation-based and program-analysis-guided random testing methods
- Create an Automotive Ethernet protection profile
- Provide a toolbox containing intelligent tools for IoT source code and online data set security analysis, as well as threat intelligence management components.
- Provide working demonstrators for automated test case generation, reliable IoT device detection and classification, as well as IoT anomaly detection and threat intelligence.
Method
Use case and research partners will iteratively design interfaces to the security assurance components and test its different modules, which the latter will develop. Using the same iterative process, automatically generated test cases, performed by program-analysis-guided random testing methods, will be integrated. Conventional anomaly detection, signal temporal logic approaches and methodologies for event correlation in CPS as well as machine learning will build the basis for the analytical toolbox. The threat intelligence management will fund on numerous mature tools, advanced standards and be shaped to the mixed ICT and industrial control system environments found in the IoT. An iterative, tool chain-based approach will be the fundament of reliable IoT device classification and network discovery at runtime, while the communication behavior in CPS will be transferred into a statistical model and builds the basis for the predictive monitoring setup for runtime trustworthiness. Power behavior of the main CPU in combination with the implemented pipeline architecture is used for the detection of power based side channel attacks.