In the current Internet era, there are now trillions of gadgets online, and the Internet of Things is becoming a necessary part of daily life. IoT devices are connected, but this also makes them vulnerable to cyberattacks. Cyberattacks targeting Internet of Things (IoT) systems have increased dramatically in both volume and sophistication in the last year. Determining the importance and explainability of significant feature selection is not done in conventional feature selection, which acts as a black box method. Classical machine learning suffers from overload and class imbalance issues in IoT-based cloud security which is the major issue that results in botnet attack detection. To overcome these two issues, this study developed a model of a feature map to encode conventional data to a quantum feature space and then utilize the newly created quantum data in the cognitive circuit, which motivates to development of Quantum Shapley Additive Explanation with Variational Quantum Classifier (QSHAP-VQC) is implemented. This makes it possible to employ classical data in a quantum circuit. To minimize a cost function, a VQC employs hybrid quantum-classical techniques that involve parameterized circuits and gates whose parameters are improved via a classically based optimization loop. A quantum-classical hybrid loop consisting of these steps is eventually broken when the classical optimization finds the ideal parameters. For training data, the usual cost function is a comparison of the actual and expected outputs. The proposed QSHAP-VQC achieves the highest rate of accuracy in the detection of attacks in an IoT cloud environment.
In the current Internet era, there are now trillions of gadgets online, and the Internet of Things is becoming a necessary part of daily life. IoT devices are connected, but this also makes them vulnerable to cyberattacks. Cyberattacks targeting Internet of Things (IoT) systems have increased dramatically in both volume and sophistication in the last year. Determining the importance and explainability of significant feature selection is not done in conventional feature selection, which acts as a black box method. Classical machine learning suffers from overload and class imbalance issues in IoT-based cloud security which is the major issue that results in botnet attack detection. To overcome these two issues, this study developed a model of a feature map to encode conventional data to a quantum feature space and then utilize the newly created quantum data in the cognitive circuit, which motivates to development of Quantum Shapley Additive Explanation with Variational Quantum Classifier (QSHAP-VQC) is implemented. This makes it possible to employ classical data in a quantum circuit. To minimize a cost function, a VQC employs hybrid quantum-classical techniques that involve parameterized circuits and gates whose parameters are improved via a classically based optimization loop. A quantum-classical hybrid loop consisting of these steps is eventually broken when the classical optimization finds the ideal parameters. For training data, the usual cost function is a comparison of the actual and expected outputs. The proposed QSHAP-VQC achieves the highest rate of accuracy in the detection of attacks in an IoT cloud environment.
Weaknesses arising from heterogeneous device structure, high-dimensional traffic patterns, and more complex nonlinear attack behaviors. The traditional intrusion detection systems (IDS), which rely on the classical machine learning models, often face the challenges of scalability, feature explosion, and low representational ability when handling large-scale IoT traffic data. To meet these challenges, we propose a Hybrid Quantum-Classical Variational Neural Network (HQCVNN) for effective intrusion detection in IoT systems. The model utilizes a Variational Quantum Circuit (VQC) as a nonlinear quantum feature map followed by dense neural layers for overall classification. Angle embedding with entanglement layers enables the encoding of quantum data and leads to a better separability of features in high-dimensional space, whilst the classical optimization guarantees efficient training on NISQ time scales. The model is implemented using Qiskit and evaluated on classical benchmark intrusion detection datasets. Experimental results show the proposed method has better classification accuracy, F1-score, and complexity reduction than conventional ANNs and SVMs. The results demonstrate the promise of hybrid quantum machine learning systems with end-to-end general-purpose capabilities in future IoT security and provide pragmatic impressions on how to deploy them on near-term quantum processors.
Abstract The rapid expansion of the Internet of Things (IoT) has produced a highly linked digital environment together with a growing number of complex cyber threats. This work presents a Quantum-Enhanced Hybrid Anomaly Detection Framework combining two advanced Quantum Machine Learning (QML) models: Quantum Support Vector Machine (QSVM) and Quantum Neural Network (QNN) to address these challenges. By projecting data into high-dimensional quantum spaces, QSVM efficiently classifies non-linear and low-frequency threats such as User to Root (U2R) and Probe attacks using quantum kernel functions. Concurrently, QNN models utilize Variational Quantum Circuits (VQCs) to detect large-scale threats like Denial of Service (DoS) and Remote to Local (R2L) attacks by offering superior generalization. Both models produce a final, high-confidence prediction by combining their outputs using a soft voting ensemble approach. The proposed framework guarantees scalability and resilience while ensuring compatibility with present simulation systems. The experimental results obtained indicate that the hybrid QSVM + QNN model significantly outperforms standalone models and conventional methods, providing a future-ready security solution for next-generation IoT infrastructures. In addition, this model delivers a robust, adaptable, and quantum-resilient security solution, well-suited for safeguarding future IoT ecosystems.
The rapid expansion of the Internet of Things in critical industrial environments has significantly increased the attack surface, exposing systems to sophisticated cyber threats. Traditional pattern-based intrusion detection systems struggle to detect such advanced attacks, while deep learning approaches, despite achieving high detection accuracy, often suffer from high computational cost and latency, limiting their deployment in resource-constrained edge gateways. Quantum machine learning offers a promising alternative by enabling high-dimensional feature representation; however, current implementations are constrained by hardware noise, limited qubit availability and backend-dependent execution characteristics in the Noisy Intermediate-Scale Quantum era. To address these challenges, this paper proposes a Residual Hybrid Quantum-Classical Neural Network (RHQ-CNN) for efficient intrusion detection in Edge-Industrial Internet of Things (Edge-IIoT) environments under idealized quantum simulation conditions. The proposed framework employs a classical neural network encoder to compress high-dimensional network traffic into a compact latent representation suitable for quantum processing. A variational quantum circuit with serial data re-uploading is then utilised to model complex non-linear decision boundaries without increasing qubit requirements. In addition, a residual connection fuses classical and quantum representations to improve training stability and preserve latent feature information. The model is evaluated on the Edge-IIoTset dataset and achieves a test accuracy of 99.94%, with high weighted performance across the 15-class detection task, although the extremely low-sample Fingerprinting class remains comparatively more challenging. Additional controlled ablation, deployment-cost, and bootstrap confidence interval analyses demonstrate the test-set metric stability and computational trade-offs of the proposed architecture. These findings highlight the potential of hybrid quantum-classical models for next-generation cybersecurity in industrial IoT systems.
This survey examines quantum-enhanced machine learning techniques for intrusion detection in Internet of Things (IoT) and fog environments, synthesizing recent work on hybrid quantum-classical models, quantum convolutional neural networks (QCNNs), quantum support vector machines (QSVMs), and quantum-assisted clustering.We review five recent primary studies that apply quantum methods to IoT security, summarize their methodological choices (encoding, circuit ansatz, classical-quantum interfaces), and evaluate reported benefits and limitations with respect to accuracy, computational cost, and deployability in resource-constrained settings.The literature indicates potential gains in feature expressivity and false-negative reduction from quantum encodings and variational circuits, but practical barriers remain: limited qubit counts, noisy hardware, encoding overhead, and scarce real-world deployment studies.We identify concrete research gaps and propose directions for lightweight encodings, noise-aware hybrid architectures, federated quantum-assisted Intrusion Detection System (IDS), and standardized benchmarks for reproducible evaluation.The survey aims to guide researchers and practitioners toward scalable, explainable, and deployable quantum-enhanced intrusion detection for IoT.
Quantum Neural Network (QNN) provides an exciting alternative to machine learning as they take advantage of quantum computer properties when making decisions and processing data. However, the integration of high-dimensional classical data with the limited number of qubits available on current Noisy Intermediate-Scale Quantum computers presents a major difficulty. This paper presents a complete hybrid quantum-classical architecture in Qiskit that utilizes the Network Security Laboratory-Knowledge Discovery and Data Mining (NSL-KDD) dataset to classify network intrusions among multiple classes. In order to work around the exponential cost of computations and reduce the number of times quantum circuits need to be run due to hardware limitations, a subset of the dataset will be created to use during both training and testing. The architecture proposed in this paper uses a four-qubit Variational Quantum Circuit (VQC). First, a classical feature vector is used to create a quantum Hilbert space via a ZZFeatureMap, which initializes the qubits via a Hadamard transformation followed by phase rotation gates that create controlled-entanglement between pairs of features. The quantum states are then passed through a RealAmplitudes variational circuit, while the classical layer optimizes the parameters using the Constrained Optimization BY Linear Approximations (COBYLA) algorithm. The overall accuracy of the model when tested on five classes of intrusion detection results in an accuracy level of 59%, with very strong classification accuracies for both Denial of Service (DoS) and User-to-Root (U2R) attacks as demonstrated by their F1 scores. The proposed QNN achieves a highly competitive baseline of 86% accuracy and 86% recall in specific configurations. These results underscore the potential of shallow, lower-complexity QNN architectures as an emergent tool for cybersecurity in the quantum era.
The rapidly emerging technology of Internet of Things (IoT) provides significant advancements in the domain of sensors and data collection in various environments, enabling improved decision making. Accuracy in data analysis, classification and fusion are raising as important challenges for the advancement of IoT. Machine Learning (ML) is expected to enhance the ability to recognize various patterns from IoT data and determine the best analytic process. These synergies are expected to have a vital role in IoT applications in industrial environment where safety and security are of utmost importance. In this work, a hybrid quantum ML classification scheme is proposed into an air quality IoT monitoring system in order to detect dangerous and abnormal situations (i.e. smoke detection and fire alarm). Furthermore, an evaluation of the proposed scheme is performed investigating the limits of classical and hybrid quantum ML approaches taking into account computational resources availability and efficiency tradeoff. In addition, it should be noted that feature reduction is also applied based on a Data Ensemble Refinement Greedy Algorithm (DERGA) optimizing the number of features employed. Finally, the noise impact over the behavior of hybrid quantum data classification is also examined in terms of accuracy under different varying parameters.