Application of Memristors in Hardware Security: A Current State‐of‐the‐Art Technology

Abstract

Memristors are widely used in hardware security applications. Research progress in memristor‐based physical unclonable functions (PUFs), random number generators (RNGs), and chaotic circuits is reviewed. To enhance device security, PUFs and RNGs apply randomness of memristors and incorporate 3D crossbars to amplify the number of challenge‐response pairs and provide proof of the destruction of the key, which enables the administrator to firmly control the device information. In addition, the image encryption technique based on the chaotic system is summarized. Furthermore, an assessment of the research advancement in PUFs, RNGs, and chaotic circuits is conducted. This Review examines the characteristics, applications, progress, and challenges of memristors in hardware security and compares the benefits and limitations of different schemes as accelerators for hardware information protection.

Secure Electronics Enabled by Atomically Thin and Photosensitive Two-Dimensional Memtransistors

“…35−38 Hybrid memristor-CMOS circuits have been used to demonstrate PUFs with higher attack resilience, lower hardware overhead, and lower energy consumption. 39,40 Similarly, hybrid memristor-CMOS circuit models have been proposed to obfuscate IC layouts using polymorphic gates and cyclic-locking schemes. 41 Models of chaotic memristive systems using Chua circuits have also been demonstrated for encryption and secure communication.…”

“…Note that several demonstrations on memristor-based crossbar arrays have utilized the process variations and inherent stochasticity for device authentication and random number generation. − Hybrid memristor-CMOS circuits have been used to demonstrate PUFs with higher attack resilience, lower hardware overhead, and lower energy consumption. , Similarly, hybrid memristor-CMOS circuit models have been proposed to obfuscate IC layouts using polymorphic gates and cyclic-locking schemes . Models of chaotic memristive systems using Chua circuits have also been demonstrated for encryption and secure communication. , While promising, two-terminal memristor-based non-von Neumann platforms are limited in computational capability and require CMOS peripherals and sensors, which increases area and energy overhead.…”

Exploiting Non-idealities of Resistive Switching Memories for Efficient Machine Learning

“…Over the last decade, the unpredictable behavior of RSMs has been demonstrated to be an efficient way to create true random number generators (Shen et al, 2021;Gaba et al, 2013;Yang et al, 2020b;Hu et al, 2016;Faria et al, 2018;Bao et al, 2020) that can be used for security purposes (Khan et al, 2021;Pang et al, 2019;Lv et al, 2020). For machine learning applications, the Gaussian nature of the stochastic distribution turns out to be an efficient way to implement probabilistic computing in hardware (Table 1).…”

Nonvolatile Resistive Switching in Layered InSe via Electrochemical Cation Diffusion

“…[9] This provides several advantages in suppressing leakage-currents and increased mechanical flexibility. [8] Recently, 2D transition-metal dichalcogenides, [10] transition-metal oxide (such as MoO 3 ), [8] hexagonal boron nitride (h-BN), [11] and elemental black phosphorus (BP) [12] have been used in vertically stacked resistive memory devices. These results are significant, however there remain quite a few areas that need further investigation in the domain of 2D material-based resistive memory technologies.…”

“…These results are significant, however there remain quite a few areas that need further investigation in the domain of 2D material-based resistive memory technologies. [5,8,[11][12][13][14][15] These include investigations into charge transport mechanisms in vertically stacked van der Waals structures and understanding the role of energy barrier of the van der Waals layers to the out-of-plane cationic and anionic diffusion.…”