IET Quantum Communication最新文献

With the continuous development of industrial intelligent technology, traditional PID algorithms have certain limitations in effectively controlling nonlinear or complex control systems. Therefore, new intelligent PID algorithms are gradually replacing traditional PID algorithms, enabling automatic parameter optimisation. However, these intelligent algorithms face challenges such as getting stuck in local optima and suboptimal convergence effects in certain complex control systems. To address these issues, this paper proposes an improved quantum genetic algorithm that encodes individuals as quantum states and applies quantum operations to update the population, obtaining classical solutions and utilising quantum crossover and quantum mutation to evolve and improve the solutions. This study conducted simulation experiments on a second-order oscillatory process using traditional PID, BP neural network (BPNN) PID, genetic algorithm (GA) BP- NN_PID and the proposed improved quantum genetic algorithm (QGA) BPNN_PID controllers, and recorded the corresponding output responses. The experimental results showed that the QGA enhanced BPNN_PID controller outperforms others, with rise time 2.8 s, overshoot 2.7% and settling time 9.4 s. Additionally, the improved controller exhibits reduced oscillation in the output and demonstrates good control performance for systems that require fast and stable output.

Quantum key distribution (QKD) allows two users to exchange a provably secure key for cryptographic applications. In prepare-and-measure QKD protocols, the states must be indistinguishable to prevent information leakage to an eavesdropper performing a side-channel attack. Here, we measure the indistinguishability of quantum states in a prepare-and-measure three-state BB84 polarisation-based decoy state protocol using resonant-cavity light-emitting diodes (RC-LEDs) as the source in the transmitter. We make the spatial, spectral and temporal DOF of the generated quantum states nearly indistinguishable using a spatial filter single-mode fibre, a narrow-band spectral filter and adjustable timing of the electrical pulses driving the RC-LEDs, respectively. The sources have fully indistinguishable transverse spatial modes. The measured fractional mutual information between an assumed eavesdropper and the legitimate receiver is $��2.39�\times �1�0^{�-�5�}�$ due to the spectral distinguishability and $��4.31�\times �1�0^{�-�5�}�$ for the temporal distinguishability. The source is integrated into a full QKD system operating in a laboratory environment, where we achieve a secure key rate of 41.5 kbits/s with an average quantum bit error rate of 2.9%. The low system size, weight and power make it suitable for mobile platforms such as uncrewed aerial vehicles (drones) or automobiles.

Medical image integrity is critical as telemedicine, cloud PACS and AI-assisted diagnostics become routine. We present a tamper localisation framework that embeds authentication signatures in the phase domain of blockwise quantum Fourier transform (QFT) coefficients. The watermark is phase-only, energy preserving and keyed through sparse midband supports with paired phase differences; a light cross-block coupling imposes spatial consistency so that localised edits produce coherent high-contrast residuals confined to manipulated regions after inverse QFT. Because magnitudes remain unaltered, benign photometric variations are naturally attenuated, improving specificity under common acquisition and storage pipelines. The verifier computes circular phase residuals and applies an adaptive threshold to generate blockwise tamper maps, which are refined to pixel resolution. Across standard distortions (JPEG recompression, Gaussian noise and blur) and localised forgeries (copy–move, inpainting and contrast edits), the scheme maintains diagnostic fidelity (typical PSNR $��>�$ 40 dB, SSIM $��>�$ 0.98) while delivering precise spatially resolved detection. The design is deterministic and reproducible via seeded keys, integrates with DICOM workflows and is amenable to future quantum hardware realisation. This work contributes a quantum-ready, imperceptible and localisation-oriented approach to medical image authentication suitable for deployment in modern healthcare systems. The proposed QFT phase–only watermark achieves imperceptibility (global PSNR $��\sim �71�$ dB; SSIM $��\sim �0.99999�$) and detects localised tampering (ROC AUC $��\sim �0.67�$ under class imbalance).

Hybrid quantum–classical architectures demonstrate significant potential for efficiently solving complex physical models. This study focuses on developing and implementing a hybrid quantum–classical algorithm to solve the Muskat–Leverett model by reformulating it as a system of linear equations. By expressing the Muskat–Leverett model in matrix form $��A�x�=�b�$, we enable the application of quantum algorithms. We leverage classical computational platforms (CPU and GPU) alongside quantum computing to efficiently solve the model, achieving competitive accuracy and demonstrating potential for scalability compared to traditional methods. The Harrow–Hassidim–Lloyd algorithm and the variational quantum linear solver were utilised to design quantum circuits and variational ansatz. The study evaluates quantum solutions with classical optimisation, variational methods, fully classical solutions on graphical processors and purely quantum solutions. Experimental results show that the hybrid architecture, utilising cuQuantum and CUDA-Q, achieves up to 6.42× faster execution compared to CPU implementations and provides a comparable speedup to GPU-based solvers. When applied for large-scale problems, the hybrid approach is also ∼10× faster than standard quantum simulators. Key factors, such as qubits, optimisers and cost functions, were analysed to determine optimal parameters. These results highlight the advantages of hybrid architecture in accelerating and optimising numerical solutions for complex physical models.

Quantum key distribution (QKD) leverages the laws of quantum mechanics to enable the secure transfer of information and is the only known method of key exchange capable of providing verifiable security. While commercial optical fibre based QKD links are available, there are presently no free space based commercial solutions. Furthermore, QKD on airborne platforms has only been experimentally verified at low altitudes and at low speeds. We first survey the academic literature to identify gaps in research and establish a foundation for our own study. Subsequently, this study utilises computational modelling to find that continuous variable (CV) QKD is feasible for high-speed airborne platforms and can provide sufficient data rates (> 100 Mb/s without multiplexing) at suitable ranges (≈15 km with full wavefront correction) for typical supersonic aircraft use cases. The implementation of such a system is economically viable as it can utilise relatively low-cost standard devices developed by the telecommunications industry, as opposed to the specialised equipment typically required for discrete variable (DV) QKD protocols. Our results highlight how the current capabilities of such devices impose limits on the performance (data rates and ranges) of airborne QKD systems.

We propose a novel procedure for generating an arbitrary energy eigenstate from physical states. To validate it, we demonstrate the generation of an energy eigenstate of the $��H_2�$ molecule from a superposition of energy eigenstates by classically emulated quantum simulation. We use a Hamiltonian represented by Pauli matrices and concatenated ancilla qubits. Starting from an adequate initial state of physical qubits, we generate a corresponding energy eigenstate by twirling operations, which evolve the system under the Hamiltonian controlled by the ancilla qubits.

Quantum technology strengthens intrusion detection with unbreakable encryption and highly precise threat sensors, providing unparallelled security against cyber threats. This work proposes the integration of quantum technology into Network Intrusion Detection Systems (IDS), focusing on Machine Learning (ML) algorithms such as Quantum Support Vector Classifier (QSVC) and Quantum Random Forest (QRF) models in comparison with Classical Support Vector Classifier (CSVC) and Classical Random Forest (CRF). The dataset used in all the models is UNSW-NB15. By harnessing vast amounts of UNSW-NB15 data, it is processed in parallel by allowing quantum technology and thereby improving computer power and speed. The results are observed simultaneously by simulating QSVC and QRF in IBM Quantum labs, which shows improved performance of $��95.3�\%�$ and $��98�\%�$ compared to classical CSVC and CRF methods, resulting in $��96�\%�$ accuracy. Later, the QSVC and QRF are processed with an optimiser to obtain improved accuracy of $��99.2�\%�$ for QSVC and $��99.78�\%�$ for QRF. The proposed framework aims to enhance the resistance and efficiency of IDS in defending against evolutionary cyber threats in a precisely interconnected digital ecosystem.

In this article, we uncover flaws and pitfalls of a quantum-based remote memory attestation procedure for Internet of things devices. We also show limitations of quantum memory that suggests the attestation problem for quantum memory is fundamentally different to the attestation problem for classical memory, even when the devices can perform quantum computation. The identified problems are of interest for quantum-based security protocol designers in general, particularly those dealing with corrupt devices. Finally, we make use of the lessons learnt to design a quantum-based attestation system for classical memory with improved communication efficiency and security.

With the rising demand for secure and reliable telecommunication networks, efficient resource allocation strategies in quantum key distribution-based software-defined optical networks (SDONs) are becoming essential. In this research, we propose a heuristic adaptive quantum routing (RWTA_AQR) algorithm for routing, wavelength and timeslot assignment. RWTA_AQR utilises the FirstFit algorithm to assign wavelengths in both contiguous and noncontiguous timeslots for optimal resource utilisation based on the priority. To verify its effectiveness, the proposed RWTA_AQR is tested on NSFNET and UBN24 network topologies under diversified traffic models, towards low as well as high demand, network congestion scenarios. We use network security performance (NSP), success ratio of connection requests, timeslot utilisation, quantum key utilisation, blocking probability (BP) and security downgrade ratio as metrics to prove its effectiveness against the existing methods based on flexible security level, strict security level and classical approach. The results demonstrate that RWTA_AQR performs better by allowing NSP of up to 90% and having the lowest BP (below 10%) at lower traffic load. The proposed solution provides a systematic trade-off in security and resource utilisation with controlled overhead for improving the performance of QKD-SDONs in dynamic resource-constrained environments.

Transforms play a pivotal role in the study of signals and images. With the advent of quantum systems, analysing signals in the quantum domain is of particular interest. In this work, we aim to build a generalised transform circuit in the quantum domain. We have shown the working of this circuit for Wavelet transform, Fourier transform, and Discrete Cosine Transform (DCT) on $��1�-�D�$ vector, $��2�-�D�$ matrix, and an image respectively. We also take inverse transforms in each of the cases to match with the given initial input. We use simulators for this work, and the results obtained are favourable. We further use this circuit for classification of heart beats as it is an essential task in detection of cardiac diseases. We utilise both classical and quantum computation to classify beats of Electrocardiogram (ECG, hereafter) signals into normal and not-normal beats (non-beats and abnormal beats). This novel architecture to carry out transform is quite general and can be used for any arbitrary transform.