[Paper Review] Detecting Recycled Commodity SoCs: Exploiting Aging-Induced SRAM PUF Unreliability

Yansong Gao, Hua Ma+3 more|arXiv (Cornell University)|May 21, 2017

Physical Unclonable Functions (PUFs) and Hardware Security3 references3 citations

TL;DR

This paper proposes a cost-free method to detect recycled commodity System-on-Chips (SoCs) by exploiting aging-induced unreliability in SRAM PUFs—hardware fingerprints derived from SRAM cell power-up states. By selecting aging-sensitive SRAM PUF responses (ASRs), the method achieves false acceptance and rejection rates below 0.001 using fewer than 1,000 responses, enabling reliable detection of field-aged SoCs without design modifications.

ABSTRACT

A physical unclonable function (PUF), analogous to a human fingerprint, has gained an enormous amount of attention from both academia and industry. SRAM PUF is among one of the popular silicon PUF constructions that exploits random initial power-up states from SRAM cells to extract hardware intrinsic secrets for identification and key generation applications. The advantage of SRAM PUFs is that they are widely embedded into commodity devices, thus such a PUF is obtained without a custom design and virtually free of implementation costs. A phenomenon known as `aging' alters the consistent reproducibility---reliability---of responses that can be extracted from a readout of a set of SRAM PUF cells. Similar to how a PUF exploits undesirable manufacturing randomness for generating a hardware intrinsic fingerprint, SRAM PUF unreliability induced by aging can be exploited to detect recycled commodity devices requiring no additional cost to the device. In this context, the SRAM PUF itself acts as an aging sensor by exploiting responses sensitive to aging. We use SRAMs available in pervasively deployed commercial off-the-shelf micro-controllers for experimental validations, which complements recent work demonstrated in FPGA platforms, and we present a simplified detection methodology along experimental results. We show that less than 1,000 SRAM responses are adequate to guarantee that both false acceptance rate and false rejection rate are no more than 0.001.

Motivation & Objective

• To address the growing threat of recycled and remarked counterfeit ICs in global supply chains, which account for over 80% of reported counterfeits.
• To extend the functionality of existing SRAM PUFs beyond authentication and key generation to include aging-based detection of recycled devices.
• To enable detection of recycled commodity SoCs without additional hardware cost or design modifications, leveraging ubiquitous SRAM PUFs already present in off-the-shelf microcontrollers.
• To develop a simplified, systematic methodology for selecting aging-sensitive PUF responses (ASRs) that maximize detection accuracy with minimal response bit length.
• To validate the method experimentally using real-world microcontrollers and demonstrate high reliability under varying aging conditions.

Proposed method

• The method identifies and selects SRAM PUF responses that exhibit high sensitivity to aging-induced reliability degradation, termed Aging-Sensitive Responses (ASRs).
• It leverages the natural aging process of SRAM cells, which alters their power-up state consistency, to serve as a built-in aging sensor.
• The approach uses statistical analysis of inter- and intra-device response differences (Δp_interA and Δp_intraA) to quantify aging effects and determine detection thresholds.
• A simplified selection strategy is proposed to prioritize ASRs with high aging sensitivity, reducing the number of required response bits for reliable detection.
• The detection process is performed by a resource-rich verifier that compares SRAM PUF responses from a device under test against reference responses from known-good devices.
• The method ensures low false acceptance (FAR) and false rejection (FRR) rates by selecting ASRs with sufficient aging-induced variance and validating performance across multiple aging periods.

Experimental results

Research questions

• RQ1 Can aging-induced unreliability in commodity SRAM PUFs be reliably exploited to detect recycled SoCs without additional hardware?
• RQ2 What is the minimum number of SRAM PUF response bits required to achieve a false acceptance rate (FAR) and false rejection rate (FRR) below 0.001?
• RQ3 How does the detection accuracy improve with increasing aging duration, and what aging period is sufficient for reliable detection?
• RQ4 Can a simplified ASR selection methodology effectively reduce the number of required response bits while maintaining high detection accuracy?
• RQ5 To what extent can existing off-the-shelf microcontrollers be used to implement this detection method without design modifications?

Key findings

• Using fewer than 1,000 SRAM PUF responses, the method achieves both false acceptance rate (FAR) and false rejection rate (FRR) below 0.001, ensuring high detection reliability.
• The required number of response bits (n) decreases by over 63% when the number of aging periods (N) increases from 3 to 9, demonstrating the effectiveness of ASR selection.
• With only nine days of field aging (equivalent to 49.6 days of accelerated aging), the method achieves EER < 10^-4 using 840 response bits for FAR and FRR below 10^-4.
• Longer aging periods (e.g., 49.6 days) reduce the number of required ASRs while maintaining the same detection accuracy, indicating improved detectability over time.
• The method is cost-free for the device, as all computations are offloaded to a verifier, and no design modifications or additional hardware are needed.
• Experimental results confirm that detection accuracy increases with prolonged aging, validating the aging-sensitive nature of the selected SRAM PUF responses.