TRNG Based on Multiple Entropy Sources Using CTDSM

True random number generators (TRNGs) extract entropy from physical processes to generate statistically independent, non-reproducible bitstreams with cryptographic-grade randomness. Silicon-based TRNGs utilize random noise in the circuit as a source of entropy to emulate an equiprobable Bernoulli random process. This paper proposes using a Continuous Time Delta-Sigma Modulator (CTDSM) to extract entropy from multiple uncorrelated noise sources for TRNG design. A CTDSM can be made entropy-rich by leveraging its negative feedback mechanism and non-idealities due to clock jitter and the quantizer’s metastability. Simulation results of the proposed design, implemented using TSMC 65 nm technology, successfully demonstrate cryptographic-grade randomness by passing all tests recommended in NIST SP 800-22.