A Spintronic SOT-MRAM Could Provide a High-Security Unclonable Challenge-Response Device for the IoT
This 1kb SOT-MRAM chip delivers 10⁹ challenge-response pairs, resists side-channel attacks, and withstands machine learning analysis.
Researchers from Beihang University, working with Truth Memory Corporation, claim to have found a new use for spintronic technology — building a 1kb spin-orbit torque magnetic memory (SOT-MRAM) that can provide physical unclonable functions (PUFs) for securing the Internet of Things (IoT).
"In recent years, physical unclonable function (PUF) has emerged as a lightweight solution in the Internet of Things security," the team explains, referring to the ability to program something into a chip that can be verified but not copied as a means of providing a security factor. "However, conventional PUFs based on complementary metal oxide semiconductor (CMOS) present challenges such as insufficient randomness, significant power and area overhead, and vulnerability to environmental factors, leading to reduced reliability."
The team's answer: spintronic technology, unsurprisingly enough given that they're from Beihang University's MIIT Key Laboratory of Spintronics. An alternative to traditional electronics, spintronics — spin electronics — is based on the effect of an electron's spin and associated magnetic moment. In this case, the spintronic device is a SOT-MRAM chip built on a 180nm CMOS process node, and used to create a 1kb physical unclonable function (PUF) security module capable of delivering 10⁹ challenge-response pairs before becoming exhausted.
"The results demonstrate that the proposed PUF exhibits near-ideal performance metrics: 50.07% uniformity, 50% diffuseness, 49.89% uniqueness, and a bit error rate of 0%, even in a 375K (101°C/215°F) environment," the team claims of the prototype. "The reconfigurability of [the] PUF is demonstrated by a reconfigurable Hamming distance of 49.31% and a correlation coefficient of less than 0.2, making it difficult to extract output keys through side-channel analysis. Furthermore, resistance to machine-learning modeling attacks is confirmed by achieving an ideal accuracy prediction of approximately 50% in the test set."
The team's work has been published in the journal Engineering, under open-access terms; no roadmap to commercialization has yet been disclosed.
