Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


Van Duc Nguyen has a total of 193 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Hybrid Geometry- and Measurement-Based UWA-MIMO Channel Modeling

    by van Duc Nguyen, Dinh Khoa Phan, Hoang Anh Nguyen, Trinh van Chien, Matthias Pätzold

    This paper proposes a double-surfaced multiple-input and multiple-output (MIMO) channel model for underwater acoustic (UWA) communications that combines geometry- and measurement-based methodologies. Fading effects and multipath propagation in shallow underwater environments are modelled by ray-tracing techniques, where the scatterers are assumed to be randomly located in seabed and seawater surface planes according to uniform distributions.We further assume that the number ofwave crests is infinite and that the waves move up and down continuously in all directions. Hence, the scatterers on the sea surface and the seabed are located on flat surfaces. Applying the geometrical channel modelling approach, we derive the impulse response of space- time-frequency selective channels for UWA MIMO communications. The spatial and temporal correlation functions are then derived from the channel impulse response. For the sake of practical applications, the frequency correlation function of the proposed channel model is matched to measurement data. The ℓp-norm method (LPNM) is applied to calculate the simulation model’s parameters. The capacity of the UWA-MIMO channel is analyzed utilizing the channel and additive ambient noise model under various practical conditions. Numerical results manifest the high accuracy of the proposed channel model and unveil the channel capacity as a multi-variate function of both the system parameters and propagation environment.

  • Top Cited Papers

    Toward Layer-Wise Personalized Federated Learning: Adaptive Layer Disentanglement via Conflicting Gradients

    by Minh-Duong Nguyen, Hoang-Khoi Do, Nam-Khanh Le, Nguyen H. Tran, Zhaohui Yang, Van-Duc Nguyen, Trinh Van Chien

    In personalized Federated Learning (pFL), high data heterogeneity can cause significant gradient divergence across devices, adversely affecting the learning process. This divergence, especially when gradients from different users form an obtuse angle during aggregation, can negate progress, leading to severe weight and gradient update degradation. To address this issue, we introduce a new approach to pFL design, namely Federated Learning with Layer-wise Aggregation via Gradient Analysis (FedLAG), utilizing the concept of gradient conflict at the layer level. Specifically, when layer-wise gradients of different clients form acute angles, those gradients align in the same direction, enabling updates across different clients toward identifying client-invariant features. Conversely, when layer-wise gradient pairs make create obtuse angles, the layers tend to focus on client-specific tasks. In hindsights, FedLAG assigns layers for personalization based on the extent of layer-wise gradient conflicts. Specifically, layers with gradient conflicts are excluded from the global aggregation process. The theoretical evaluation demonstrates that when integrated into other pFL baselines, FedLAG enhances pFL performance by a certain margin. Therefore, our proposed method achieves superior convergence behavior compared with other baselines. Extensive experiments show that our FedLAG outperforms several state-of-the-art methods and can be easily incorporated with many existing methods to further enhance performance.


  • Open Access

    Single- and Multi-Objective Stochastic Optimization for Next-Generation Networks in the Generative AI and Quantum Computing Era

    by Trinh Van Chien, Bui Trong Duc, Nguyen Xuan Tung, Van Duc Nguyen, Waqas Khalid, Symeon Chatzinotas, Lajos Hanzo

    Next Generation (NG) networks move beyond simply connecting devices to creating an ecosystem of connected intelligence, especially with the support of generative Artificial Intelligence (AI) and quantum computation. These systems are expected to handle large-scale deployments and high-density networks with diverse functionalities. As a result, there is an increasing demand for efficient and intelligent algorithms that can operate under uncertainty from both propagation environments and networking systems. Traditional optimization methods often depend on accurate theoretical models of data transmission, but in real-world NG scenarios, they suffer from high computational complexity in large-scale settings. Stochastic Optimization (SO) algorithms, designed to accommodate extremely high density and extensive network scalability, have emerged as a powerful solution for optimizing wireless networks. This includes various categories that range from model-based approaches to learning-based approaches. These techniques are capable of converging within a feasible time frame while addressing complex, large-scale optimization problems. However, there is currently limited research on SO applied for NG networks, especially the upcoming Sixth-Generation (6G). In this survey, we emphasize the relationship between NG systems and SO by eight open questions involving the background, key features, and lesson learned. Overall, our study starts by providing a detailed overview of both areas, covering fundamental and widely used SO techniques, spanning from single to multi-objective signal processing. Next, we explore how different algorithms can solve NG challenges, such as load balancing, optimizing energy efficiency, improving spectral efficiency, or handling multiple performance trade-offs. Lastly, we highlight the challenges in the current research and propose new directions for future studies.

  • Energy-Efficient and Real-Time Sensing for Federated Continual Learning via Sample-Driven Control

    by Minh Ngoc Luu, Minh-Duong Nguyen, Ebrahim Bedeer, Van Duc Nguyen, Dinh Thai Hoang, Diep N. Nguyen, Quoc-Viet Pham

    An intelligent Real-Time Sensing (RTS) system must continuously acquire, update, integrate, and apply knowledge to adapt to real-world dynamics. Managing distributed intelligence in this context requires Federated Continual Learning (FCL). However, effectively capturing the diverse characteristics of RTS data in FCL systems poses significant challenges, including severely impacting computational and communication resources, escalating energy costs, and ultimately degrading overall system performance. To overcome these challenges, we investigate how the data distribution shift from ideal to practical RTS scenarios affects Artificial Intelligence (AI) model performance by leveraging the generalization gap concept. In this way, we can analyze how sampling time in RTS correlates with the decline in AI performance, computation cost, and communication efficiency. Based on this observation, we develop a novel Sample-driven Control for Federated Continual Learning (SCFL) technique, specifically designed for mobile edge networks with RTS capabilities. In particular, SCFL is an optimization problem that harnesses the sampling process to concurrently minimize the generalization gap and improve overall accuracy while upholding the energy efficiency of the FCL framework. To solve the highly complex and time-varying optimization problem, we introduce a new soft actor-critic algorithm with explicit and implicit constraints (A2C-EI). Our empirical experiments reveal that we can achieve higher efficiency compared to other DRL baselines. Notably, SCFL can significantly reduce energy consumption up to 85% while maintaining FL convergence and timely data transmission.


  • Open Access

    Deep Learning-based Prediction of Alertness and Drowsiness using EEG Signals

    by Cong Thuan Do, Anh Nam Le, Van Duc Nguyen, Van Chien Trinh

    In this paper, we propose to apply two advanced deep learning models, comprising Bidirectional long short-term memory (LSTM) and Transformer, to predict the alertness or drowsiness of individuals via utilizing electroencephalogram (EEG) signals. In the both models, we extract some features from brainwave data including alpha, beta, delta, and theta waves. For the Bidirectional LSTM model, we construct a four-layer deep learning network to learn brainwave features and classify the human statuses. For the Transformer model, we use an encoder and replace the position encoding with the vector representation of time to capture good features of sequential brainwave data. Experimental results demonstrate that the two proposed models provide very high accuracy, up to 99.55, in predicting the alertness and drowsiness of individuals.


  • Open Access

    Proposal for a Doppler Shift Compensation Method Using Non-Uniform FFT with Pilot Carrier Frequency for OFDM-Based Underwater Acoustic Communication Systems

    by Van Duc Nguyen, Thi Hoai Linh Nguyen, Quoc Khuong Nguyen, Tien Hoa Nguyen

    In this paper, we propose a method that uses the non-uniform Fast Fourier Transform (FFT) to compensate for Doppler frequency shifts in OFDM-based underwater acoustic (UWA) communication systems. To estimate the Doppler frequency shift, the paper proposes using a pilot carrier frequency (PCF) that is identified by transmitting with greater power in the frequency domain compared to other subcarriers. This identifier helps estimate both the Doppler frequency shift and the channel. The paper proposes estimating and compensating for the Doppler frequency shift in two steps: Step 1 performs coarse frequency synchronization, where the Doppler frequency shift is determined by using PCF signal. The Doppler shift correction is obtained by re-sampling the received signal, and then interpolating re-sampled signal. However, the Doppler shift compensation in the step 1 cannot correct the phase distortion of the measured PCF signal. Therefore, performing step 2, known as fine frequency synchronization, is necessary. In this step, the correction for the phase distortion is determined based on the phase difference between the two consecutive OFDM signals in one frame. The remaining Doppler frequency shift is adjusted based on the phase deviation, using the non-uniform FFT. By using the non-uniform FFT, the complexity of the ICI compensation is significantly reduced, and the quality of Doppler shift compensation is improved. The experience results show that the transmitted text will be decoded correctly by using the proposed technique.

  • Open Access

    Deep Learning-Assisted Channel Estimation in Frequency Selective UWA Communication Systems

    by Van Duc Nguyen, Dinh Khoa Phan, Trinh Van Chien

    Envisioned future wireless gives rise to the possibility of exploiting underwater applications with high spectral efficiency and reliable communication. This paper presents a method to combine channel estimation in the time domain with neural networks and supervised learning to improve orthogonal frequency-division multiplexing (OFDM) systems in underwater communications. The pilot structure is designed in such a way that the propagation channels are tracked from symbol to symbol. Moreover, the low-density parity-check code (LDPC) channel coding method is applied to overcome the severe fading and attenuation effects in underwater acoustic environments. We consider this proposed system’s performance in a measurement-based channel model to evaluate it. In particular, the system’s performance is evaluated for different levels of channel mobility. Numerical results show that, with the assistance of deep learning, the channel estimation performance can be improved, depending on how fast the channel changes. Compared with state-of-the-art benchmarks, our proposal offers better system performance in terms of bit error ratio and normalized mean square error.