Privacy-Aware Framework of Robust Malware Detection in Indoor Robots: Hybrid Quantum Computing and Deep Neural Networks

TL;DR

The paper addresses DoS- and spoofing-related threats to indoor robotic CPS by proposing a privacy-aware, hybrid quantum–classical malware detection framework that integrates quantum feature encoding with dropout-optimized DNNs. It demonstrates a modular QNN–DNN pipeline that can run with limited qubits on NISQ hardware and uses privacy-preserving telemetry, QuXAI overlays, and confidence-weighted fusion to maintain interpretability and safety. Empirically, the approach achieves up to $95.238%$ detection accuracy and $F1>0.95$ even under privacy constraints, generalizes robustly, and shows resilience to training instability. The work advances trustworthy AI for CPS by delivering scalable, privacy-aware, and explainable malware detection suitable for real-time indoor robotics deployments.

Abstract

Indoor robotic systems within Cyber-Physical Systems (CPS) are increasingly exposed to Denial of Service (DoS) attacks that compromise localization, control and telemetry integrity. We propose a privacy-aware malware detection framework for indoor robotic systems, which leverages hybrid quantum computing and deep neural networks to counter DoS threats in CPS, while preserving privacy information. By integrating quantum-enhanced feature encoding with dropout-optimized deep learning, our architecture achieves up to 95.2% detection accuracy under privacy-constrained conditions. The system operates without handcrafted thresholds or persistent beacon data, enabling scalable deployment in adversarial environments. Benchmarking reveals robust generalization, interpretability and resilience against training instability through modular circuit design. This work advances trustworthy AI for secure, autonomous CPS operations.

Figures (16)

• Figure 1: Hybrid quantum-classical architecture with staggered layout.
• Figure 2: Quantum neural network architecture with input, hidden, and output layers. Each arrow represents a quantum perceptron unitary $U_j^l$ acting between layers.
• Figure 3: Quantum backpropagation flow. Forward propagation of $\rho^{\text{in}}$ through CP maps $\xi^l$, and backward propagation of the target state via adjoint channels $\mathcal{F}^l$. Dashed arrows indicate layer-wise gradient computation.
• Figure 4: Spoof-Aware Annotation: Spoof-aware annotation showing adversarial injection points in the system.
• Figure 5: NISQ-compatible hybrid quantum malware detection pipeline. Input features are encoded into quantum states, processed by a VQC, and optimized via classical feedback. Outputs are fused with classical logic for robust and interpretable detection.