Biological physically unclonable function

Akshay Wali,Sahin Kaya Ozdemir,Saptarshi Das,Andrew Pannone,Likhith Kumar Reddy Usthili,Yang Wu,Ibrahim Tarik Ozbolat,Akhil Dodda

doi:10.1038/s42005-019-0139-3

Akshay Wali, Sahin Kaya Ozdemir + Show 6 more

Open Access

https://doi.org/10.1038/s42005-019-0139-3

Copy DOI

Abstract

Information security is one of the foundational requirements for any modern society thriving on digital connectivity. At present, information security is accomplished either through software algorithms or hardware protocols. Software algorithms use pseudo random numbers generated by one-way mathematical functions that are computationally robust in the classical era, but are shown to become vulnerable in the post-quantum era. Hardware security overcomes such limitations through physically unclonable functions (PUFs) that exploit manufacturing process variations in the physical microstructures of Si integrated circuits to obtain true random numbers. However, recent upsurge in reverse engineering strategies make Si-PUFs vulnerable to various attacks. Moreover, Si-PUFs are low-entropy, power-hungry, and area-inefficient. Here we introduce a biological PUF which exploits the inherent randomness found in the colonized populations of T cells and is difficult to reverse engineer and at the same time is high-entropy, non-volatile, reconfigurable, ultra-low-power, low-cost, and environment friendly.

Highlights

Information security is one of the foundational requirements for any modern society thriving on digital connectivity
physically unclonable functions (PUFs) are associated with the inherent randomness in the physical microstructures of the hardware, mostly in integrated circuits (ICs), which can address the shortcomings of one-way mathematical functions
Various types of Si-PUFs16 have been proposed: arbiter PUFs are based on random variations in interconnect and/or transistor delays, whereas, static random access memory (SRAM) and dynamic random access memory (DRAM) based PUFs exploit variation in transistor dimensions manifesting in the fluctuation in the ON state current and gate delays

Summary

Introduction

Information security is one of the foundational requirements for any modern society thriving on digital connectivity. Software algorithms use pseudo random numbers generated by one-way mathematical functions that are computationally robust in the classical era, but are shown to become vulnerable in the post-quantum era Hardware security overcomes such limitations through physically unclonable functions (PUFs) that exploit manufacturing process variations in the physical microstructures of Si integrated circuits to obtain true random numbers. The recent decline of the Si technology owing to the slowdown of Moore’s law of scaling and the emergence of dark Si have influenced and facilitated the growth in IoT technology based on novel nanomaterials and devices[28,29] In this context, nanotech PUFs based on randomly dispersed nanoparticles[30,31,32], self-assembled carbon nanotubes (CNTs)[33], sub-lithographic random network of metal wires, block copolymers, and memristive crossbar arrays[34,35,36] are being investigated. It is possible in the near future to integrate some of these benign bio-species with smart devices for on-chip information security opening opportunities for innovation

Methods

Results

Conclusion