Non-unitary Quantum Physical Unclonable Functions Model Open System Dynamics, Resisting Machine Learning and Side-channel Attacks

Physical Unclonable Functions (PUFs) offer a promising route to hardware security by harnessing inherent physical randomness, but their classical designs are increasingly vulnerable to sophisticated attacks. Mohammadreza Vali, Hossein Aghababa, and Nasser Yazdani, from the University of Tehran and Loyola University Maryland, address this challenge by exploring the potential of quantum mechanics to create more robust PUFs. Their research introduces a new class of non-unitary QPUFs that move beyond idealised models and actively incorporate the effects of environmental noise, such as decoherence and dissipation, as a fundamental security feature. The team proposes three distinct architectures, the Dissipative, Measurement-Feedback, and Lindbladian QPUFs, and demonstrates through simulation that these designs achieve strong performance in terms of uniqueness, uniformity, and resistance to forgery, offering a pathway towards post-quantum hardware authentication and a new understanding of how to leverage noise for security.

Researchers delve into the theoretical underpinnings, practical implementation, and potential applications of QPUFs, including secure device authentication, robust key generation, and secure communication protocols. QPUFs leverage the principles of quantum mechanics, such as superposition, entanglement, and measurement, to enhance security and resilience.

The no-cloning theorem, a fundamental principle of quantum mechanics, underpins the potential advantage of QPUFs over classical approaches. Researchers present QPUFs as a means to generate truly random keys and securely authenticate devices, with randomness stemming from the inherent unpredictability of quantum measurements. The team explores techniques to mitigate the susceptibility of quantum systems to noise and decoherence, including error correction and careful system design. Scientists utilize Quantum State Tomography to characterize quantum systems and assess the quality of generated randomness.

The SWAP test, a quantum algorithm, measures the similarity between quantum states, evaluating the uniqueness of QPUF responses. The Lindblad Master Equation, a mathematical framework, models the evolution of open quantum systems, accounting for noise and decoherence. Unitary t-Designs ensure the randomness and unpredictability of QPUF responses, while Quantum Key Distribution serves as a complementary technology for secure key exchange. Researchers employ Von Neumann Measurement, a standard quantum measurement technique, and utilize GHZ and Cluster States, specific types of entangled quantum states.

This research highlights the potential of QPUFs in diverse applications, including secure device authentication, protecting vulnerable IoT devices, securing critical infrastructure, and verifying the integrity of remote memory. Researchers propose novel approaches such as Biometric Feature-Dimension Cryptography, combining biometric data with quantum principles, and Noise-Assisted Quantum Autoencoders, improving quantum machine learning algorithms. The team also explores Quantum Broadcasting for secure information distribution, Quantum Secret Sharing for collaborative secret reconstruction, and Hybrid Authentication Protocols combining QPUFs with other security mechanisms. Researchers addressed the vulnerability of existing QPUFs to advanced machine learning and side-channel attacks by harnessing non-unitary effects, specifically decoherence and dissipation, as a foundation for unforgeable hardware authentication. The study pioneers three distinct QPUF architectures, each embedding physical irreversibility into the challenge-response process. The Dissipative QPUF utilizes amplitude damping as an entropy source, introducing randomness through the loss of quantum information to the environment.

The Measurement-Feedback QPUF employs mid-circuit measurements coupled with conditional unitary operations, creating a feedback loop sensitive to environmental interactions. Most comprehensively, the Lindbladian QPUF models Markovian noise via the Lindblad master equation, a mathematical framework describing the evolution of open quantum systems. This model incorporates a decomposition to efficiently simulate the complex interactions between the system and its environment. Extensive simulations assessed the trade-offs between entropy generation, reproducibility, and hardware feasibility, revealing practical limitations and opportunities for optimization.

The L-QPUF demonstrated exponential modeling resistance under limited challenge-response access, signifying a substantial improvement in security against sophisticated attacks. Researchers quantified the similarity and distinguishability of quantum states using Uhlmann fidelity and trace distance, while the diamond distance compared noisy channels to their ideal unitary counterparts. Researchers demonstrate that by harnessing non-unitary dynamics, the natural evolution of quantum systems interacting with their environment, it is possible to create inherently unclonable devices. Three distinct QPUF architectures were proposed and evaluated through detailed simulations. The Dissipative QPUF utilizes amplitude damping, a process of natural quantum decay, as its primary entropy source, effectively exploiting unavoidable quantum loss as a security feature.

The Measurement-Feedback QPUF integrates mid-circuit measurements with classical feedback, creating stochastic evolution paths uniquely determined by the device’s history and the measurement process. Most significantly, the Lindblad QPUF employs a generalized framework based on open quantum system theory, modelling device behaviour using Lindblad master equations and their channel decompositions. Simulations demonstrate that these non-unitary designs maintain high levels of uniqueness, uniformity, and unforgeability, while allowing for controlled trade-offs in reliability due to the stochastic nature of the quantum channels. Crucially, the L-QPUF architecture achieves exponential modelling resistance under limited challenge-response access, indicating a substantial improvement in security against attempts to clone or predict the device’s responses. These findings highlight that non-unitary evolution can serve as a practical foundation for post-quantum hardware authentication, offering a pathway toward scalable, inherently unclonable quantum devices. The research establishes a theoretical and computational framework for noise-aware quantum hardware authentication, demonstrating that environmental noise can be reframed as a constructive security resource.

Noise-Driven QPUF Designs for Authentication

This work introduces a new approach to quantum physical unclonable functions (QPUFs), moving beyond traditional designs that rely on ideal quantum operations. Researchers developed and analysed three distinct architectures, the Dissipative QPUF, Measurement-Feedback QPUF, and Lindblad QPUF, each leveraging different aspects of quantum noise to create unique and secure device fingerprints. The Dissipative QPUF utilizes amplitude damping, a natural process of quantum decay, as its primary source of entropy, effectively exploiting unavoidable quantum loss as a security feature. The Measurement-Feedback QPUF integrates mid-circuit measurements with classical feedback, creating stochastic evolution paths uniquely determined by the device’s history and the measurement process.

The Lindblad QPUF employs a generalized framework based on open quantum system theory, modelling device behaviour using Lindblad master equations and their channel decompositions. Simulations demonstrate that these non-unitary designs maintain high levels of uniqueness, uniformity, and unforgeability, while allowing for controlled trade-offs in reliability due to the stochastic nature of the quantum channels. The Lindblad.