(A) study of self-clocking true random number generator based on a stochastic oscillation by a single MOSFET

Abstract

The true random number generator (TRNG) is demonstrated by using intrinsic stochasticity of the oscillation of a single transistor, essential for non-decryptable hardware-based security devices. Here, we propose a novel and advanced true random number generator utilizing stochastic oscillation behavior of a single transistor exhibiting evident advantages in randomness, scalability, circuit complexity, power consumption. In this work, a single transistor true random number generator is studied by using the single transistor latch (STL) phenomenon in bipolar junction transistor (BJT) operation. The randomness od this device exhibits advantages on previously suggested non-volatile characteristics based True random number generators. And this device’s low power characteristic is advantageous compared to previously reported volatile characteristics-based true random number generators by using analog-to-digital converter to generate random bits. The random number generating operation is shown in both silicon-on-insulator (SOI) and bulk silicon substrate. By using physical floating body, stochastic oscillation is operated on SOI substrate. But it has disadvantages on cost using SOI substrate compare to bulk silicon substrate. The operating principle of stochastic oscillation is based on electrical floating body by biasing voltage on substrate or using deep n-well under p-well body. True random number generator on bulk silicon substrate has advantages on higher scalability and lower cost compare to using SOI substrate. The STL and stochastic oscillation of poly-Si channel MOSFET is also demonstrated so that the possibility of monolithic 3D stackable RNG is suggested. Improved energy efficiency compared to SOI MOSFET is shown thanks to irregular grain boundaries of poly-Si channel.