Enhancing Physical Unclonable Function Robustness Employing Embedded Instruments

This paper proposes a methodology to improve the robustness of CMOS physical unclonable functions with regard to environmental parameter variations and thus enhancing the security. The methodology exploits the reuse of embedded instruments which are being deployed for dependability purposes. The approach is hardware and power efficient which is especially important for applications such as IoT. Chip implementation of the embedded instruments in 40nm CMOS technology is presented in addition to simulations and experimental validation on an FPGA.