Highly secure cryptography is a topic of increasing concern, especially in the digital age with globalization in an unprecedented scale. Moreover, with internet of things growing very fast a completely new huge market has been opened for secure communication. This is a topic of great importance especially when it comes to autonomous driving where manipulated communication could result in fatalities. As transport infrastructure is built to long-term, there are also high demands for cryptography as well as all other utilized technology. However, in order to guarantee a long-term high level of security, current asymmetric cryptographic algorithms are not suitable.
Furthermore, the long-term security evaluation of methods for key distribution, data integrity and authenticity poses the main problem. Current method are most namely Diffie-Hellman key exchange, hybrid key distribution techniques and digital signature (RSA, DSA, ECDSA). However, all of the before named method’s safety is derived from assumptions that cannot be proven for sure (complexity of prime factorization or dis¬crete logarithm). In addition, quantum computers are a serious threat towards those methods and thus should not be used in the distant future anymore. As a result, physical methods and post-quantum systems are topic of growing importance within cryptography. Although quantum cryptography is an excellent replacement ensuring the necessary security, it is very expensive and not suitable for autonomous driving due to the vast amount of objects. For the last couple of years research has been concerned with new physical methods for generating and distributing cryptographic keys by utilizing high-frequency radio communication. It is relying on the reciprocity of the communication channel. This approach also comes with drawbacks, especially on the topics of a reasonably high key entropy, eavesdropping, object-authenticity and communications over a bigger distance. Hence, the research project KIF is targeting based on the current state-of-the-art, these specific issues.
Moreover, two challenges, which have not yet been discussed a lot, are to be handled. On the one hand there is the fast generation of highly secure cryptographic keys with a big amount of different objects as it occurs with autonomous driving. The other big challenge that is identified is the communication over bigger distances. For this second part methods need to be researched as reciprocity cannot be taken for granted anymore. Thus, the key generation would here be based on a good connection between the communication partners as well as pattern detection.
The project KIF researches highly secure methods to generate and distribute cryptographic keys utilizing channel data and in addition other data enabling even less powerful hardware to be more resilient to attacks by quantum computers. The methods should ensure sufficient entropy for the key, object authenticity, data integrity and eavesdropping protection while being applicable for short-range as well as long-range communications. Furthermore, this project aims to fulfill full functionality for autonomous driving with challenges such as high object velocities, fast changing objects, a large amount of different objects, rough environment and later on a small and cheap implementation.