Deep neural networks (DNNs) are phenomenally successful machine learning methods broadly applied to many different disciplines. However, as complex two-party computations, DNN inference using classical cryptographic methods cannot achieve unconditional security, raising concern on security risks of DNNs' application to sensitive data in many domains. We overcome such a weakness by introducing a quantum-aided security approach. We build a quantum scheme for unconditionally secure DNN inference based on quantum oblivious transfer with an untrusted third party. Leveraging DNN's noise tolerance, our approach enables complex DNN inference on comparatively low-fidelity quantum systems with limited quantum capacity. We validated our method using various applications with a five-bit real quantum computer and a quantum simulator. Both theoretical analyses and experimental results demonstrate that our approach manages to operate on existing quantum computers and achieve unconditional security with a negligible accuracy loss. This may open up new possibilities of quantum security methods for deep learning.
Deep neural networks (DNNs) are phenomenally successful machine learning methods broadly applied to many different disciplines. However, as complex two-party computations, DNN inference using classical cryptographic methods cannot achieve unconditional security, raising concern on security risks of DNNs' application to sensitive data in many domains. We overcome such a weakness by introducing a quantum-aided security approach. We build a quantum scheme for unconditionally secure DNN inference based on quantum oblivious transfer with an untrusted third party. Leveraging DNN's noise tolerance, our approach enables complex DNN inference on comparatively low-fidelity quantum systems with limited quantum capacity. We validated our method using various applications with a five-bit real quantum computer and a quantum simulator. Both theoretical analyses and experimental results demonstrate that our approach manages to operate on existing quantum computers and achieve unconditional security with a negligible accuracy loss. This may open up new possibilities of quantum security methods for deep learning.
Deep learning in quantum computing seeks to leverage the unique properties of quantum systems, such as superposition and entanglement, to enhance the performance of deep learning algorithms. Quantum neural networks (QNNs), which are designed to operate on quantum computers, have the potential to enable faster and more efficient inference execution. However, quantum computers are susceptible to noise, which can rapidly degrade the coherence of quantum states and lead to errors in quantum computations. As a result, deep neural networks (DNNs) that operate on quantum computers may experience degraded classification accuracy during inference. However, in this paper, we demonstrate that this intrinsic quantum noise can actually improve the robustness of DNNs against adversarial input attacks. The noisy behavior of quantum computers can reduce the impact of adversarial attacks, thereby improving the accuracy of the degraded DNNs. To further enhance DNN robustness, we perform am extensive exploration on the prowess of Quantum Noise injection for Adversarial Defense (QNAD), which induces carefully crafted crosstalk in the quantum computer. QNAD preselects a subset of pretrained network weights to be perturbed with injected crosstalk in the qubits, causing them to become entangled due to interactions between neighboring qubits. When evaluated on state-of-the-art network dataset configurations, the proposed QNAD approach provides up to 268% relative improvement in accuracy, against adversarial input attacks compared to conventional DNN implementations.
Quantum computing represents a major threat to future communication systems security: Once a certain level of maturity is reached, the security enjoyed by classical cryptographic schemes will immediately be put in check by (i) highly efficient quantum algorithms able to solve hard problems non-negligibly faster than the best-known classical solutions and (ii) potentially scalable fabrication techniques for quantum circuits. This work investigates the design of quantum-safe security primitives suitable for new cybersecurity threats in next-generation communications networks by means of the integrated use of quantum cryptography and deep learning. By combining the unconditionally secure guarantees offered by quantum cryptographic protocols with the adaptive, data-driven intelligence of deep learning models, this work proposes a holistic solution to secure communication infrastructures against not only current but also future cyber threats. A central element is the utilization of QKD-protocols that use the properties of quantum mechanics in order to securely distribute keys that are resistant to computational attacks. In parallel thereof, deep learning techniques are applied to monitor the system by analyzing real-time network traffic, identifying abnormal activities, and forecasting sophisticated attack patterns that frequently are overlooked by conventional security systems. This approach mitigates constraints of standalone quantum cryptography, including noise and hardware imperfections, and improves threat identification over traditional signature-based systems. The work studies the synergy between quantum cryptography and AI-assisted analytics, optimizing quantum channel estimation, error correction, and intrusion detection with deep learning. We conduct extensive simulations to show the efficacy of the framework against various adversarial attacks such as man in the middle, side-channel, and adversarial machine learning threats. The system has high stability without loss of quantum bit error rate, and high detection accuracy, for handling the change of attack vectors in distributed, heterogeneous communication networks.
Secure multiparty computation enables the joint evaluation of multivariate functions across distributed users while ensuring the privacy of their local inputs. This field has become urgent due to the demand for computationally intensive deep learning inference. These computations are typically offloaded to cloud servers, leading to vulnerabilities. To solve this problem, we introduce a linear algebra engine that leverages the quantum nature of light for information-theoretically secure multiparty inference using telecommunication components. We apply this linear algebra engine to deep learning and derive rigorous upper bounds on the information leakage of both the deep neural network weights and the client’s data, enabling double-blind operations. Applied to the modified National Institute of Standards and Technology classification task, we obtain test accuracies exceeding 95% while guaranteeing leakage of less than 0.1 bits per weight and data element. This leakage is an order of magnitude below the minimum bit precision required for accurate deep learning using state-of-the-art quantization techniques. Our work lays the foundation for practical quantum-secure computation and unlocks secure cloud deep learning as a field.
Generation of secure communication was getting known as a promising technology called as a quantum cryptography. Unfortunately, it still faces several challenges, most notably in terms of high computational demands, scalability limitations, quantum decoherence, and vulnerability to side channel attacks, such that deployment of it in real world remains impossible yet. This study shows how Machine Learning (ML) and Deep Learning (DL) can be applied to tune the mentioned obstacles so that quantum cryptographic frameworks become more secure, more efficient and more scalable. In particular, we introduce a hybrid AI enabled model where RL can be utilized for tuning the performance of the post quantum cryptographic algorithm implementations, GANs can be adopted for measuring the robustness of the system, and FL can be used to make the quantum key distribution scalable. Besides, this thesis uses Convolutional Neural Networks (CNNs) and Recurrent Neural Networks (RNNs) to incorporate the quantum authentication and key exchange methods. Additionally, our methodology employs techniques based on Graph Neural Networks (GNNs) to achieve the best performance in the networks and Adversarial Machine Learning (AML) to counter and detect, and then reduce cyber threats in the run time. For preferable cryptographic computations, we introduce Quantum Neural Networks (QNNs) to reduce its dependency on expensive quantum hardware. Experimental results also show that ML/DL based quantum security frameworks provide lowering of compute burden, improvement in real time security of data, and improve resistance to cyber threats. The result of this roadmap based on AI driven strategy is a comprehensive view, with a high level of direction describing the main steps to lead to post quantum security and take benefit of quantum cryptography and apply it in a multitude of significant utilization including secure communication, cryptographic financial systems or national infrastructure protection.
Quantum machine learning (QML) is promising for potential speedups and improvements in conventional machine learning (ML) tasks (e.g., classification/regression). The search for ideal QML models is an active research field. This includes identification of efficient classical-to-quantum data encoding scheme, construction of parametric quantum circuits (PQC) with optimal expressivity and entanglement capability, and efficient output decoding scheme to minimize the required number of measurements, to name a few. However, most of the empirical/numerical studies lack a clear path towards scalability. Any potential benefit observed in a simulated environment may diminish in practical applications due to the limitations of noisy quantum hardware (e.g., under decoherence, gate-errors, and crosstalk). We present a scalable quantum-classical hybrid deep neural network (DeepQMLP) architecture inspired by classical deep neural network architectures. In DeepQMLP, stacked shallow Quantum Neural Network (QNN) models mimic the hidden layers of a classical feed-forward multi-layer perceptron network. Each QNN layer produces a new and potentially rich representation of the input data for the next layer. This new representation can be tuned by the parameters of the circuit. Shallow QNN models experience less decoherence, gate errors, etc. which make them (and the network) more resilient to quantum noise. We present numerical studies on a variety of classification problems to show the trainability of DeepQMLP. We also show that DeepQMLP performs reasonably well on unseen data and exhibits greater resilience to noise over QNN models that use a deep quantum circuit. DeepQMLP provided up to 25.3% lower loss and 7.92% higher accuracy during inference under noise than QMLP.
Quantum embedding is a fundamental prerequisite for applying quantum machine learning techniques to classical data and has substantial impacts on performance outcomes. In this study, we present neural quantum embedding (NQE), a method that efficiently optimizes quantum embedding beyond the limitations of positive and trace-preserving maps by leveraging classical deep-learning techniques. NQE enhances the lower bound of the empirical risk, leading to substantial improvements in classification performance. Moreover, NQE improves robustness against noise. To validate the effectiveness of NQE, we conduct experiments on IBM quantum devices for image data classification, resulting in a remarkable accuracy enhancement from 0.52 to 0.96. In addition, numerical analyses highlight that NQE simultaneously improves the trainability and generalization performance of quantum neural networks, as well as of the quantum kernel method.