This work presents an over-the-air (OTA) OFDM software-defined radio (SDR) prototype using GNU Radio and two HackRF One devices to evaluate classical and quantumkernel channel estimation.LS, LMMSE, Support Vector Regression (SVR), and quantum-kernel SVR (QSVR) are compared on identical received IQ samples under controlled Gaussian and impulsive noise.BER, packet recovery, latency, and processing throughput are evaluated.Results show that the kernel-based estimators remain competitive with conventional approaches, demonstrating the feasibility of quantum-kernel channel estimation in a complete over-the-air SDR receiver.
This work presents an over-the-air (OTA) OFDM software-defined radio (SDR) prototype using GNU Radio and two HackRF One devices to evaluate classical and quantumkernel channel estimation.LS, LMMSE, Support Vector Regression (SVR), and quantum-kernel SVR (QSVR) are compared on identical received IQ samples under controlled Gaussian and impulsive noise.BER, packet recovery, latency, and processing throughput are evaluated.Results show that the kernel-based estimators remain competitive with conventional approaches, demonstrating the feasibility of quantum-kernel channel estimation in a complete over-the-air SDR receiver.
Channel estimation in OFDM systems requires minimal complexity with one-tap equalizers and it is generally performed based on pilot symbols (PS) using least squares (LS). However, in a practical environment where impulsive noise may be present, this method may not be effective; furthermore, the minimal complexity required depends on cyclic prefixes (CP), which must be sufficiently large to cover the channel impulse response. In contrast, the use of PS and CP decreases the useful information that can be conveyed in an OFDM frame, thereby degrading the spectral efficiency of the system. In this context, we propose the use of quantum machine learning algorithms for channel estimation in OFDM systems with a reduced number of PS, without PS, and without CP. Our approach involves adapting classical machine learning models, specifically support vector regression and K-means clustering, to their quantum counterparts by integrating quantum kernels and encoding strategies suitable for channel modeling. The performance of the resulting quantum models is evaluated in comparison to LS and classical learning-based estimators. The viability of our approach is substantiated by computational simulation results obtained in frequency-selective channel models with the presence of non-Gaussian impulsive noise interfering with the symbols.
Abstract This paper presents a novel hybrid architecture that integrates Software Defined Radio (SDR) with Quantum Machine Learning (QML) for secure, adaptive, and high-performance wireless communication. The proposed system utilizes Offset Quadrature Amplitude Modulation (OQAM) for enhanced spectral efficiency and reduced inter-symbol interference, benefiting from the nonlinear signal properties managed by filter bank-based multicarrier systems. Quantum Support Vector Machines (QSVM), Quantum Reservoir Computing (QRC), and Quantum Neural Networks (QNN) are deployed to classify modulation types and manage spectrum allocation in real-time. Simulation results derived from GNU Radio and the PennyLane framework indicate that the suggested system attains a modulation classification accuracy of up to 96.3%. The QNN model trained on OQAM signals achieves an accuracy of 94.7% and an end-to-end processing latency of around 16.2 milliseconds. The design reliably sustains classification performance over 90% throughout a signal-to-noise ratio (SNR) range of 0 to 20 dB, even in the presence of noise. Secure transmission is guaranteed by a conventional control interface that enables Quantum Key Distribution (QKD), included inside the SDR environment. The integration of reconfigurable SDR hardware with nonlinear quantum machine learning algorithms positions this method as a formidable option for dynamic spectrum optimization and secure communications in forthcoming 5G and 6G networks.
Quantum Support Vector Regression (QSVR) has emerged as one of the most promising quantum machine learning algorithms for nonlinear regression on Noisy Intermediate-Scale Quantum (NISQ) devices. However, despite numerous simulation-based studies, experimental evaluations on physical quantum hardware remain relatively scarce. This work presents a comparative analysis of three regression approaches: classical Support Vector Regression (SVR) with a Radial Basis Function (RBF) kernel, classical SVR utilizing precomputed quantum kernels, and the native QSVR implementation available in the qiskit-machine-learning framework. Experiments were conducted on three benchmark regression datasets using three execution environments: a noiseless Qiskit Aer simulator, the IQM Adonis hardware-aware simulator, and the physical 5-qubit IQM Spark superconducting quantum computer (Odra 5) located at Wrocław University of Science and Technology. Multiple quantum feature maps and circuit depths were investigated, while model hyperparameters were optimized using grid search. The obtained results demonstrate that QSVR can be successfully executed on contemporary superconducting quantum hardware and, for selected datasets, achieves prediction accuracy comparable to or exceeding that of a classical RBF-based SVR. The experiments further reveal the significant influence of quantum hardware noise, feature map selection, and circuit depth on regression quality. Moreover, the implementation based on precomputed quantum kernels generally exhibits greater robustness on physical hardware than the native QSVR implementation. The presented benchmark provides an experimental assessment of QSVR under realistic execution conditions and contributes to the ongoing evaluation of quantum machine learning algorithms on current NISQ-era quantum computers.
Abstract Quantum machine learning is a rapidly growing field at the intersection of quantum computing and machine learning. In this work, we examine our quantum machine learning models, which are based on quantum support vector classification (QSVC) and quantum support vector regression (QSVR). We investigate these models using a quantum-circuit simulator, both with and without noise, as well as the IonQ Harmony quantum processor. For the QSVC tasks, we use a dataset containing fraudulent credit card transactions and image datasets (the MNIST and the Fashion-MNIST datasets); for the QSVR tasks, we use a financial dataset and a materials dataset. For the classification tasks, the performance of our QSVC models using 4 qubits of the trapped-ion quantum computer was comparable to that obtained from noiseless quantum-circuit simulations. The result is consistent with the analysis of our device-noise simulations with varying qubit-gate error rates. For the regression tasks, applying a low-rank approximation to the noisy quantum kernel, in combination with hyperparameter tuning in ε-SVR, improved the performance of the QSVR models on the near-term quantum device. Our results suggest that the quantum kernel, as described by our shallow quantum circuit, can be effectively used for both QSVC and QSVR tasks, indicating its resistance to noise and its adaptability to various datasets.
This paper examines the application of a Quantum Support Vector Machine (QSVM) for radarbased aerial target classification using micro-Doppler signatures. Classical features are extracted and reduced via Principal Component Analysis (PCA) to enable efficient quantum encoding. The reduced feature vectors are embedded into a quantum kernel-induced feature space using a fully entangled ZZFeatureMap and classified using a kernel based QSVM. Performance is first evaluated on a quantum simulator and subsequently validated on NISQ-era superconducting quantum hardware, specifically the IBM Torino (133-qubit) and IBM Fez (156-qubit) processors. Experimental results demonstrate that the QSVM achieves competitive classification performance relative to classical SVM baselines while operating on substantially reduced feature dimensionality. Hardware experiments reveal the impact of noise and decoherence and measurement shot count on quantum kernel estimation, and further show improved stability and fidelity on newer Heron r2 architecture. This study provides a systematic comparison between simulator-based and hardware-based QSVM implementations and highlights both the feasibility and current limitations of deploying quantum kernel methods for practical radar signal classification tasks.
Quantum Support Vector Regression (QSVR) has become an exciting prospect of modeling the nonlinear relationship of high-dimensional data by using a kernel method. Nevertheless, the overall assessment of the whole QSVR pipeline, including preprocessing and dimensionality reduction, quantum embedding, shot-based kernel estimation, and noise-aware analysis, is not widespread. This paper introduces a working Python implementation of the QSVR framework that combines a principal component analysis (PCA)-to-qubit embedding, a 1024-shot quantum kernel implementation, and 1024-shot 1/2-SVR optimization on financial time-series and credit transaction regression problems. Classical features are normalized and minimized to around 20 major features that capture 95% of the variance in the data before being angle-based RY embedded using a depth-3 entangling circuit. Quantum kernels are built through state-overlap estimation on finite-shot constraints, and they are trained by cross-validated optimization. The robustness of the proposed framework is measured in both noise-free and simulator-generated noise conditions in case of controlled noise approximations. Experimentation indicates improved regression performance than that of an RBF-SVR baseline with the same preprocessing and validation conditions. The suggested workflow designates the concept of reproducibility, benchmarking control, and feasible noise-conscious evaluation, which offers a systematic reference structure to subsequent research within near-term quantum machine learning.