DECIFR: Domain-Aware Exfiltration of Circuit Information from Federated Gradient Reconstruction

Federated Learning (FL) is a promising approach for multiparty collaboration as a privacy-preserving technique in hardware assurance, but its security against adversaries with domain-specific knowledge is underexplored. This paper demonstrates a critical vulnerability where available standard cell library layouts (SCLL) can be exploited to compromise the privacy of sensitive integrated circuit (IC) training data. We introduce DECIFR, a novel two-stage Membership Inference Attack (MIA) that requires no auxiliary dataset. The attack employs a guided Gradient Inversion Attack (GIA) to reconstruct a client's training images from intercepted model updates. Our findings reveal that the fidelity of these reconstructions directly correlates with membership status, allowing an adversary to reliably distinguish members from non-members based on image quality. This work exposes a practical threat that overcomes the limitations of conventional attacks and underscores that standard FL protocols are insufficient for securing domains with extensive knowledge. We conclude that robust defenses are essential for the secure application of FL in hardware assurance.

Key Contributions

First data-free MIA for FL segmentation models using standard cell library layouts to guide gradient inversion without auxiliary datasets
Demonstrates that reconstruction fidelity correlates with membership status, enabling reliable member/non-member distinction
Shows adversaries can infer hardware-specific traits (layers, technology nodes) by exploiting domain knowledge in FL hardware assurance

🛡️ Threat Analysis

Model Inversion Attack

Uses gradient inversion attack to reconstruct client training images from intercepted model updates in federated learning — adversary reconstructs private IC training data from gradients.

Membership Inference Attack

Primary contribution is a membership inference attack that determines if specific IC images were in the training set by analyzing reconstruction fidelity.