Emerging memristive devices have been recently suggested for use in secret key generation and other hardware security applications. This position paper brings together the views of researchers from material science and hardware-oriented security. It discusses the question which types of memristors are better suitable for the construction of major hardware security primitives. Specifically, this paper points out the problems caused by electroforming, a necessary step for most of today's memristive devices, and advocates the usage of electroforming-free memristors. It discusses which security properties can be met by such devices and where more research is required.
展开▼
