An Asynchronous and Low-Power True Random Number Generator Using STT-MTJ

<div>The Spin Transfer Torque Magnetic Tunnel Junction</div><div>(STT-MTJ) is an emerging memory technology whose interesting</div><div>stochastic behavior might benefit security applications. In this</div><div>paper, we leverage this stochastic behavior to construct a true</div><div>random number generator (TRNG), the basic module in the</div><div>process of encryption key generation. Our proposed TRNG</div><div>operates asynchronously and thus can use small and fast STT</div><div>MTJ devices. As such, it can be embedded in low-power and</div><div>low-frequency devices without loss of entropy. We evaluate</div><div>the proposed TRNG using a numerical simulation, solving the</div><div>Landau–Lifshitz–Gilbert (LLG) equation system of the STTMTJ</div><div>devices. Design considerations, attack analysis, and process</div><div>variation are discussed and evaluated. The evaluation shows that</div><div>our solution is robust to process variation, achieving a Shannonentropy</div><div>generating rate between 99.7Mbps and 127.8Mbps for</div><div>90% of the instances.</div>