The Total Resilience research group pursues a dual approach. First, concrete concepts for the enhancement of the resilience of systems are investigated, e.g. for improved IT security solutions and reliability concepts. This includes approaches for a resilient design of hardware accelerators for cryptographic procedures, concepts for the resilient design of sensor nodes and communication protocols which are respectively implemented and analyzed. These rather application-oriented aspects are at least partially supported by other departments. The second scientific focus lies on the research of exemplified metrics for the evaluation of resilience. This includes testing of the accomplished degree of resilience as well as methodologies for the development of resilient systems.
In the frame of theoretical considerations with regard to the understanding of resilience as a holistic system concept, the integration of the cognition aspect as an essential asset for the treatment of unknown fault situations has earned international recognition as the brand essence. The definition of resilience measures was successfully advanced and will be the basis for the evaluation of resilience mechanisms in the frame of further investigations. A further essential aspect is the development of theoretical models for the prediction of resilience characteristics in development processes. Here, the challenge is that the holistic approach for the resilience of material characteristics, the design of ASICs up to communication protocols must be understood and modelled. As already outlined, the development of side-channel-attack-resistant crypto-accelerators is extremely challenging and complex. This demand grows by a further order of magnitude for the development of resilient systems. The investigations in the area of crypto-accelerators have shown that system characteristics such as resilience against side-channel-attacks not only depend on the used hardware platform for realization, but also on the operating conditions. This means that during the development of resilient systems the operating conditions and additionally their changes in the life cycle of the equipment covering several years must be taken into account. Such a prediction of the operating parameters and the implementation of all possible safety measures cannot be completely realized at the time of design. Therefore it follows, as already mentioned, that systems must possess cognition. This is the foundation for the introduction of appropriate measures. As a mutual vertical application topic in the frame of the matrix structure of the group, e-Health systems were identified.
- design of resilient CPSoS e.g. in the field of e-Health
- development of a design methodology for resilient systems
- hardware security
- artificial intelligence
- resilience: metrics, methods, specific solutions and design methodologies
- Side Channel Attacks: analysis and countermeasures
- efficient implementations of cryptographic algorithms
- signal processing and AI based data processing
- implementation of AI accelerators
- use of methods of AI for Side Channel Attacks (Best Paper Award PIMRC Workshops - W4, 2019)
- Side Channel Attacks against commercially available authentification chips (Best Paper Award SecHard Workshop, 2019)
- SECI – Security-Interpreter-Tools for platform independent realization of intrusion detection systems (safety solutions) on sensor nodes (supported operating systems RIOT, LangOS (IHP), Linux, Window) (ACM MobiHoc, Workshops)