Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Log inSign up
Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Blog
Job Board
Q3AS

© 2026 Aqora Quantum S.A.S.

TermsPrivacyLegal Notice

arXiv:2609.04271 · IEEE Transactions on Network Science and Engineering

Quantum-Assisted Memory-Efficient Training for Parameter-Intensive Wi-Fi-Based Human Activity Recognition

To Truong An, Jie Zhang, Guolin Yin, Junqing Zhang, Yanjiao Li, Trung Q. Duong, Simon L. Cotton

Published Jan 1, 2026• 0 citations
arXivHTMLPDF

Abstract

Wi-Fi-based human activity recognition (HAR) has become an important part of integrated sensing and communications, paving the way for a range of context-aware services. However, most existing Wi-Fi-based HAR systems rely on deep learning (DL) models that are computationally and memory intensive in both training and inference, which poses significant challenges for real-world deployment. Conventional training requires simultaneous updates of millions of parameters, leading to prohibitive memory consumption. In this paper, we propose a novel quantum-assisted memory-efficient training framework (Q-MET) designed to improve efficiency in both training and inference. Q-MET utilizes a hybrid quantum classical neural network to indirectly generate parameters for HAR models, significantly reducing the trainable parameter count compared to direct optimization. To further support the deployment on resource-constrained devices, we integrate structured pruning during the training phase. Experimental results demonstrate that Q-MET achieves a 90% to 95% reduction in trainable parameters compared with conventional backpropagation-based DL training while maintaining or even exceeding classical classification accuracy. Additionally, Q-MET supports lightweight inference through structured pruning, achieving 75% to 85% model sparsity with less than 2% loss in classification accuracy. To the best of our knowledge, this work represents the first quantum-assisted approach to simultaneously tackle memory inefficiencies in both the training and inference stages of HAR systems.

DOI
10.1109/tnse.2026.3725484

Related papers

Towards Quantum Efficient Training for Radio Frequency Fingerprint Identification

To Truong An, Guolin Yin, Junqing Zhang +3 more

Radio frequency fingerprint identification (RFFI) is an emerging physical-layer method for authenticating devices through their unique hardware impairments. Deep learning (DL) is widely used to identify devices based on their signal transmissions. However, training DL models is resource-intensive because it requires repeated updates to numerous parameters, which becomes particularly problematic in resource-constrained environments. In this paper, we propose quantum-assisted training (QAST), a framework that addresses training inefficiencies in RFFI systems. QAST integrates a quantum neural network with a classical neural network to generate parameters for a DL model. Compared to traditional training methods, this indirect training strategy substantially decreases the number of trainable parameters, thereby mitigating the overall computational and resource demands. Experimental results show that QAST enables training an RFFI model with 90% fewer trainable parameters than traditional training approaches, while maintaining comparable classification accuracy.

IEEE International Conference on Communications workshops•May 24, 2026•0 citations

A Quantum-Optimized Training Framework for Radio Frequency Fingerprint Identification

To Truong An, Guolin Yin, Junqing Zhang +3 more

Radio frequency fingerprint identification (RFFI) offers a promising physical-layer method to authenticate devices based on unique hardware impairments. However, existing RFFI systems use deep learning (DL) models that are resource- intensive. Training is particularly demanding, requiring repeated updates to a large number of parameters. In this paper, we introduce a quantum-assisted training (QAST) framework to address training inefficiencies in RFFI. QAST integrates a quantum neural network (QNN) with a mapping network to generate parameters for a classical DL model. This indirect training strategy substantially reduces the number of trainable parameters and the overall memory requirements compared to direct training of the DL model. We achieve this by introducing a multimodal mapping network that effectively learns the QNN output. This network generates multiple classical parameters from a shared quantum representation, thereby reducing qubit requirements and lowering the risk of barren-plateau-related trainability degradation. We also propose a new embedding method that reduces the size of the embedding matrix and yields a 15% to 30% reduction in training time. The tailored QAST framework trains the RFFI model while requiring only 10% of the original number of parameters while maintaining comparable accuracy, thereby substantially reducing memory and computational overhead and enabling efficient training or retraining in resource-limited environments.

IEEE Journal on Selected Areas in Communications•Jan 1, 2026•0 citations

AutoQML: Automated Quantum Machine Learning for Wi-Fi Integrated Sensing and Communications

Toshiaki Koike–Akino, Pu Perry Wang, Ye Wang

Commercial Wi-Fi devices can be used for integrated sensing and communications (ISAC) to jointly exchange data and monitor indoor environment. In this paper, we investigate a proof-of-concept approach using automated quantum machine learning (AutoQML) framework called AutoAnsatz to recognize human gesture. We address how to efficiently design quantum circuits to configure quantum neural networks (QNN). The effectiveness of AutoQML is validated by an in-house experiment for human pose recognition, achieving state-of-the-art performance greater than 80% accuracy for a limited data size with a significantly small number of trainable parameters.

arXiv:2205.09115•IEEE Sensor Array and Multichannel Signal Processing Workshop (SAM)•Jun 20, 2022•19 citations

Federated Quantum Kernel-Based Long Short-Term Memory for Human Activity Recognition

Yu-Chao Hsu, Jiun-Cheng Jiang, Chun-Hua Lin +5 more

In this work, we introduce the Federated Quantum Kernel-Based Long Short-term Memory (Fed-QK-LSTM) framework, integrating the quantum kernel methods and Long Shortterm Memory into federated learning. Within Fed-QK-LSTM framework, we enhance human activity recognition (HAR) in privacy-sensitive environments and leverage quantum computing for distributed learning systems. The DeepConv-QK-LSTM architecture on each client node employs convolutional layers for efficient local pattern capture, this design enables the use of a shallow QK-LSTM to model long-range relationships within the HAR data. The quantum kernel method enables the model to capture complex non-linear relationships in multivariate time-series data with fewer trainable parameters. Experimental results on RealWorld HAR dataset demonstrate that Fed-QK-LSTM framework achieves competitive accuracy across different client settings and local training rounds. We showcase the potential of Fed-QK-LSTM framework for robust and privacy-preserving human activity recognition in real-world applications, especially in edge computing environments and on scarce quantum devices.

arXiv:2508.06078•IEEE International Conference on Quantum Computing and Engineering (QCE)•Aug 30, 2025•3 citations

Quantum-Enhanced Learning Framework for Intelligent and AI-Native 6G Wireless Networks

Muhammad Bilal Akram Dastagir, Rakesh Saini, Omer Tariq +4 more

The recent convergence of 6G wireless systems and Tiny Machine Learning (TinyML) has driven the need for on-device intelligence in edge networks, where ultra-low latency, stringent energy budgets, and tight compute constraints demand novel architectures. Lightweight deep models efficiently extract local patterns but fail to capture global dependencies, while attention mechanisms do so at the expense of energy and computational cost. To bridge this gap, we introduce Quantumer, a hybrid TinyML--quantum framework that integrates multi-scale dilated convolutions and scaled dot-product attention within a lightweight transformer architecture, employing a two-stage transfer learning pipeline from Quantum Pre-Training (Quantumer-Q) to Classical Fine-Tuning (Quantumer-C). We also present QuantiblentLayer, a four-qubit variational circuit that maps compact traffic representations into measurement-based Hilbert-space features using trainable rotations and cyclic entangling operations. The circuit is used only during offline pre-training as a nonlinear embedding teacher and is removed before Quantumer-C deployment, leaving a fully classical inference model without runtime quantum execution. By transferring these quantum-assisted embeddings into an energy-efficient, lightweight transformer, Quantumer achieves strong detection performance with minimal compute and memory overhead on resource-constrained edge devices. The intrusion detection system (IDS) is used as a case study and evaluated on the Edge-IIoTset, TON IoT, and WUSTL-IIoT-2021 datasets. Quantumer-Q achieves competitive compact-model performance with 105.86K parameters, 0.4038 MB memory usage, 0.5525 MB model size, and 5.5646 MFLOPs; the INT8 Raspberry Pi 4 deployment obtains 16.8413 ms latency with a 0.6493 MB footprint. These results support training-time quantum-assisted representation learning for compact edge-deployable IDS.

quant-pharXiv:2609.06814•Sep 8, 2026•0 citations

Quantum-Enhanced Neural Architectures for Real-Time Contextual NLP in Resource-Constrained Environments

Aakansha Soy, Kokhkhorova Madinakhon Abdukodir Kizi, K.S. Bhuvaneshwari +3 more

Among existing technologies, quantum computing presents great potential in natural language processing (NLP), especially in resource-poor environments where the method of deep learning is hampered by the related computational burden. It introduces a Quantum Assisted Hybrid Neural Network (QHNN) framework by combining various quantum enhanced algorithms with very lightweight neural architectures for real-time, minimal resource-consuming contextual NLP processing. For dimensionality reduction, efficient reduction, the proposed architecture uses the Quantum Variational Autoencoders (QVAE), for contextual understanding, Quantum Attention Mechanism (QAM) and for better sequence modeling, Quantum Kernelized LSTM (QK-LSTM). We also present Quantum Aware Federated Learning (QAFL) to perform the distributed model training across the low power edge devices to provide data privacy. Finally, it is evaluated in terms of efficiency, accuracy, and consumption of energy compared to traditional deep learning models on simulated quantum circuits and benchmarked NLP datasets. Results show that QHNN achieves great reductions in computational complexity with high contextual accuracy, and is thus suitable for edge AI application including IoT conversational agent, low latency chatbots, and mobile real time language translation. Quantum-assisted NLP is this study's contribution to the gap in demonstrating the feasibility of quantum computing advancements for practical deployment of quantum computing for real-world AI applications. Recently, the autonomy has also been extended to this problem by hybrid quantum classical models that when combined with near-term quantum hardware still provide significant scaling and adaptability.

International Conference on Innovations and Emerging Technologies in AI & Communication Systems (IETACS)•Nov 6, 2025•0 citations
Wi-Fi-based human activity recognition (HAR) has become an important part of integrated sensing and communications, paving the way for a range of context-aware services. However, most existing Wi-Fi-based HAR systems rely on deep learning (DL) models that are computationally and memory intensive in both training and inference, which poses significant challenges for real-world deployment. Conventional training requires simultaneous updates of millions of parameters, leading to prohibitive memory consumption. In this paper, we propose a novel quantum-assisted memory-efficient training framework (Q-MET) designed to improve efficiency in both training and inference. Q-MET utilizes a hybrid quantum classical neural network to indirectly generate parameters for HAR models, significantly reducing the trainable parameter count compared to direct optimization. To further support the deployment on resource-constrained devices, we integrate structured pruning during the training phase. Experimental results demonstrate that Q-MET achieves a 90% to 95% reduction in trainable parameters compared with conventional backpropagation-based DL training while maintaining or even exceeding classical classification accuracy. Additionally, Q-MET supports lightweight inference through structured pruning, achieving 75% to 85% model sparsity with less than 2% loss in classification accuracy. To the best of our knowledge, this work represents the first quantum-assisted approach to simultaneously tackle memory inefficiencies in both the training and inference stages of HAR systems.
Radio frequency fingerprint identification (RFFI) is an emerging physical-layer method for authenticating devices through their unique hardware impairments. Deep learning (DL) is widely used to identify devices based on their signal transmissions. However, training DL models is resource-intensive because it requires repeated updates to numerous parameters, which becomes particularly problematic in resource-constrained environments. In this paper, we propose quantum-assisted training (QAST), a framework that addresses training inefficiencies in RFFI systems. QAST integrates a quantum neural network with a classical neural network to generate parameters for a DL model. Compared to traditional training methods, this indirect training strategy substantially decreases the number of trainable parameters, thereby mitigating the overall computational and resource demands. Experimental results show that QAST enables training an RFFI model with 90% fewer trainable parameters than traditional training approaches, while maintaining comparable classification accuracy.
Radio frequency fingerprint identification (RFFI) offers a promising physical-layer method to authenticate devices based on unique hardware impairments. However, existing RFFI systems use deep learning (DL) models that are resource- intensive. Training is particularly demanding, requiring repeated updates to a large number of parameters. In this paper, we introduce a quantum-assisted training (QAST) framework to address training inefficiencies in RFFI. QAST integrates a quantum neural network (QNN) with a mapping network to generate parameters for a classical DL model. This indirect training strategy substantially reduces the number of trainable parameters and the overall memory requirements compared to direct training of the DL model. We achieve this by introducing a multimodal mapping network that effectively learns the QNN output. This network generates multiple classical parameters from a shared quantum representation, thereby reducing qubit requirements and lowering the risk of barren-plateau-related trainability degradation. We also propose a new embedding method that reduces the size of the embedding matrix and yields a 15% to 30% reduction in training time. The tailored QAST framework trains the RFFI model while requiring only 10% of the original number of parameters while maintaining comparable accuracy, thereby substantially reducing memory and computational overhead and enabling efficient training or retraining in resource-limited environments.
Commercial Wi-Fi devices can be used for integrated sensing and communications (ISAC) to jointly exchange data and monitor indoor environment. In this paper, we investigate a proof-of-concept approach using automated quantum machine learning (AutoQML) framework called AutoAnsatz to recognize human gesture. We address how to efficiently design quantum circuits to configure quantum neural networks (QNN). The effectiveness of AutoQML is validated by an in-house experiment for human pose recognition, achieving state-of-the-art performance greater than 80% accuracy for a limited data size with a significantly small number of trainable parameters.
In this work, we introduce the Federated Quantum Kernel-Based Long Short-term Memory (Fed-QK-LSTM) framework, integrating the quantum kernel methods and Long Shortterm Memory into federated learning. Within Fed-QK-LSTM framework, we enhance human activity recognition (HAR) in privacy-sensitive environments and leverage quantum computing for distributed learning systems. The DeepConv-QK-LSTM architecture on each client node employs convolutional layers for efficient local pattern capture, this design enables the use of a shallow QK-LSTM to model long-range relationships within the HAR data. The quantum kernel method enables the model to capture complex non-linear relationships in multivariate time-series data with fewer trainable parameters. Experimental results on RealWorld HAR dataset demonstrate that Fed-QK-LSTM framework achieves competitive accuracy across different client settings and local training rounds. We showcase the potential of Fed-QK-LSTM framework for robust and privacy-preserving human activity recognition in real-world applications, especially in edge computing environments and on scarce quantum devices.
The recent convergence of 6G wireless systems and Tiny Machine Learning (TinyML) has driven the need for on-device intelligence in edge networks, where ultra-low latency, stringent energy budgets, and tight compute constraints demand novel architectures. Lightweight deep models efficiently extract local patterns but fail to capture global dependencies, while attention mechanisms do so at the expense of energy and computational cost. To bridge this gap, we introduce Quantumer, a hybrid TinyML--quantum framework that integrates multi-scale dilated convolutions and scaled dot-product attention within a lightweight transformer architecture, employing a two-stage transfer learning pipeline from Quantum Pre-Training (Quantumer-Q) to Classical Fine-Tuning (Quantumer-C). We also present QuantiblentLayer, a four-qubit variational circuit that maps compact traffic representations into measurement-based Hilbert-space features using trainable rotations and cyclic entangling operations. The circuit is used only during offline pre-training as a nonlinear embedding teacher and is removed before Quantumer-C deployment, leaving a fully classical inference model without runtime quantum execution. By transferring these quantum-assisted embeddings into an energy-efficient, lightweight transformer, Quantumer achieves strong detection performance with minimal compute and memory overhead on resource-constrained edge devices. The intrusion detection system (IDS) is used as a case study and evaluated on the Edge-IIoTset, TON IoT, and WUSTL-IIoT-2021 datasets. Quantumer-Q achieves competitive compact-model performance with 105.86K parameters, 0.4038 MB memory usage, 0.5525 MB model size, and 5.5646 MFLOPs; the INT8 Raspberry Pi 4 deployment obtains 16.8413 ms latency with a 0.6493 MB footprint. These results support training-time quantum-assisted representation learning for compact edge-deployable IDS.
Among existing technologies, quantum computing presents great potential in natural language processing (NLP), especially in resource-poor environments where the method of deep learning is hampered by the related computational burden. It introduces a Quantum Assisted Hybrid Neural Network (QHNN) framework by combining various quantum enhanced algorithms with very lightweight neural architectures for real-time, minimal resource-consuming contextual NLP processing. For dimensionality reduction, efficient reduction, the proposed architecture uses the Quantum Variational Autoencoders (QVAE), for contextual understanding, Quantum Attention Mechanism (QAM) and for better sequence modeling, Quantum Kernelized LSTM (QK-LSTM). We also present Quantum Aware Federated Learning (QAFL) to perform the distributed model training across the low power edge devices to provide data privacy. Finally, it is evaluated in terms of efficiency, accuracy, and consumption of energy compared to traditional deep learning models on simulated quantum circuits and benchmarked NLP datasets. Results show that QHNN achieves great reductions in computational complexity with high contextual accuracy, and is thus suitable for edge AI application including IoT conversational agent, low latency chatbots, and mobile real time language translation. Quantum-assisted NLP is this study's contribution to the gap in demonstrating the feasibility of quantum computing advancements for practical deployment of quantum computing for real-world AI applications. Recently, the autonomy has also been extended to this problem by hybrid quantum classical models that when combined with near-term quantum hardware still provide significant scaling and adaptability.