True random number generation using voltage controlled spin-dice

Abhronil Sengupta, Akhilesh Jaiswal, Kaushik Roy

Research output: Chapter in Book/Report/Conference proceedingConference contribution

4 Scopus citations

Abstract

True Random Number Generators (TRNGs) are becoming increasingly popular in cryptography and other security applications. However, conventional TRNG designs in hardware often result in significantly high area and power consumption [1] and hence recent research efforts have been directed to developing compact, low power and high throughput TRNGs based on emerging technologies like the Magnetic Tunnel Junction (MTJ 'spin-dice') [2]. The random number generation process usually takes place through the application of two current pulses, namely the 'reset' pulse to orient the magnet to a known initial state and subsequently the 'roll' pulse to switch the magnet with probability of 0.5. The stochastic switching nature of the MTJ arises from the inherent thermal noise present in the device. However, the quality of the random number generated is not sufficiently high due to variations in the magnitude of current required to switch the MTJ with 50% probability (arising from PVT variations). Hence expensive post-processing schemes are usually required [2]. In this work, we explore the design of a Voltage Controlled Spin-Dice (VC-SD) using the recently discovered phenomena of Voltage Controlled Magnetic Anisotropy (VCMA) in an MTJ structure to orient the ferromagnet along a meta-stable magnetization direction and subsequently utilizing thermal noise to produce random switching of the magnet to either one of the stable magnetization directions. In addition to power and reliability benefits, the proposed TRNG is able to provide better resiliency against PVT variations.

Original languageEnglish (US)
Title of host publication74th Annual Device Research Conference, DRC 2016
PublisherInstitute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)9781509028276
DOIs
StatePublished - Aug 22 2016
Event74th Annual Device Research Conference, DRC 2016 - Newark, United States
Duration: Jun 19 2016Jun 22 2016

Publication series

NameDevice Research Conference - Conference Digest, DRC
Volume2016-August
ISSN (Print)1548-3770

Other

Other74th Annual Device Research Conference, DRC 2016
CountryUnited States
CityNewark
Period6/19/166/22/16

All Science Journal Classification (ASJC) codes

  • Electrical and Electronic Engineering

Fingerprint Dive into the research topics of 'True random number generation using voltage controlled spin-dice'. Together they form a unique fingerprint.

Cite this