True Random Number Generation in Nonlinear Internal-Resonating MEMS Resonators

This work demonstrates for the first time a true random number generator (TRNG) based on an internal resonating (IR) micromechanical resonator. In particular, the resonator launches IR and exhibits quasi-periodic motions once the driving frequency passes a clear transition boundary defined by the saddle-node on an invariant circle (SNIC) bifurcation. By employing an open-loop driving configuration with a fixed frequency near the boundary, the fluctuation of the thermal noise in the background as the entropy source causes a shift in the bifurcation boundary, leading to significant changes in the system dynamics, and in turn randomly generated IR induced frequency combs. The presence or absence of frequency sidebands serves as the basis for bits of “1” or “0”, respectively. Here, a total of 102,400 random bits captured passes the National Institute of Standards and Technology (NIST) statistical test without post-processing, offering a promising approach for generating true random numbers for hardware encryption especially in extreme applications where pure electronic devices fail to survive.