Hardware-Rooted PUF Fingerprinting for Device-Level Traceability in Knowledge Distillation

2026-09-25

AI Summary

This study addresses the challenge of tracing compromised devices following model intellectual property theft in knowledge distillation. We propose a cross-architecture identity inheritance mechanism that embeds hardware Physical Unclonable Function (PUF) fingerprints into teacher logits. By leveraging Ring Oscillator PUFs implemented on Xilinx FPGAs to generate hardware-rooted fingerprints, our approach achieves device-level identity transfer across heterogeneous networks through two-stage neural decoding and Hamming distance correction. As the first hardware-rooted provenance scheme for this context, it demonstrates robustness against post-processing tampering while maintaining high detection accuracy under noisy conditions. Ultimately, this work provides a practical tracking system for distributed AI deployments, effectively mitigating the risks associated with unauthorized model extraction and redistribution.

Abstract

Knowledge distillation (KD) enables model transfer across heterogeneous platforms and deployment environments, yet it exposes proprietary models to distillation-based theft, where an adversary extracts intellectual property (IP) by training a student model on teacher outputs. Current defenses, such as software watermarking or hardware-based access control, either fail to survive the distillation process or lack the granularity to identify the specific device responsible for a leak. Aiming to provide post-theft accountability and traceability, we propose a novel fingerprinting framework that superimposes device-specific Physical Unclonable Function (PUF) signatures onto teacher logits during distillation. By utilizing signatures derived from Ring Oscillator (RO) PUFs measured on a Xilinx Zynq-7020 FPGA, we ensure that any student model trained via KD inherits a unique, hardware-linked identity. Our framework is architecture-agnostic, enabling reliable identity inheritance across heterogeneous structures, including Convolutional Neural Networks, Vision Transformers, and encoders. To ensure robust attribution, we implement a two-stage recovery pipeline consisting of a neural decoder and Hamming-distance refinement, maintaining high detection accuracy even under noisy conditions. Furthermore, we introduce a multi-level logit encoding scheme to support large-scale device deployment. Experimental results demonstrate that the embedded fingerprints are resilient against common post-distillation modifications. These results establish a practical system-level approach for enabling hardware-linked model traceability in distributed AI deployment environments.

Problem

• Knowledge Distillation
• Model Theft
• Device-Level Traceability
• Intellectual Property Protection
• Physical Unclonable Function

Innovation

• Physical Unclonable Function (PUF)
• Knowledge Distillation
• Hardware-Rooted Fingerprinting
• Logit Encoding
• Model Traceability