Novel Randomized Biased Placement for FPGA Based Robust Random Number Generator with Enhanced Uniqueness

摘要

The physical unclonable functions (PUF) have widely been used to provide software as well as hardware security for the cyber-physical systems. They are used for performing significant cryptography tasks such as generating keys, device authentication and securing against IP piracy. They have also been used to produce the root of trust. However, they lack in reliability metric. We present a novel approach for improving the uniqueness as well as the reliability of field programmable gated arrays (FPGAs) based ring oscillator PUF and derive a random number, consuming a very small area concerning look up tables (LUTs). We use profiling method for observing frequency variations in ring oscillators (RO), spatially placed across the FPGA floor, and are able to spot the suitable locations for RO mapping, which leads to enhanced ROPUF reliability. We have implemented proposed methodology on Xilinx -7 series FPGAs and tested the robustness against environmental variations e.g. temperature and supply voltage variations. The proposed approach achieves 6% higher uniqueness of 49.83% along with the reliability of 99.35%, which as a group of PUF characteristics, is a significant improvement as compared to characteristics provided by existing ROPUF methods. The random number generator so realized passes all applicable nine tests of NIST uniformity statistical test suite.