arXiv subjects

Dongwei Xu

Publications and source records attributed to Dongwei Xu.

At least 19 recordsLinked to original sources

arXiv2025

Improving the Convergence Rate of Ray Search Optimization for Query-Efficient Hard-Label Attacks

In hard-label black-box adversarial attacks, where only the top-1 predicted label is accessible, the prohibitive query complexity poses a major obstacle to practical deployment. In this paper, we focus on optimizing a representative class of attacks that search for the optimal ray direction yielding the minimum $\ell_2$-norm perturbation required to move a benign image into the adversarial region. Inspired by Nesterov's Accelerated Gradient (NAG), we propose a momentum-based algorithm, ARS-OPT, which proactively estimates the gradient with respect to a future ray direction inferred from accumulated momentum. We provide a theoretical analysis of its convergence behavior, showing that ARS-OPT enables more accurate directional updates and achieves faster, more stable optimization. To further accelerate convergence, we incorporate surrogate-model priors into ARS-OPT's gradient estimation, resulting in PARS-OPT with enhanced performance. The superiority of our approach is supported by theoretical guarantees under standard assumptions. Extensive experiments on ImageNet and CIFAR-10 demonstrate that our method surpasses 13 state-of-the-art approaches in query efficiency.

cs.LG↗

arXiv2025

\emph{FoQuS}: A Forgetting-Quality Coreset Selection Framework for Automatic Modulation Recognition

Deep learning-based Automatic Modulation Recognition (AMR) model has made significant progress with the support of large-scale labeled data. However, when developing new models or performing hyperparameter tuning, the time and energy consumption associated with repeated training using massive amounts of data are often unbearable. To address the above challenges, we propose \emph{FoQuS}, which approximates the effect of full training by selecting a coreset from the original dataset, thereby significantly reducing training overhead. Specifically, \emph{FoQuS} records the prediction trajectory of each sample during full-dataset training and constructs three importance metrics based on training dynamics. Experiments show that \emph{FoQuS} can maintain high recognition accuracy and good cross-architecture generalization on multiple AMR datasets using only 1\%-30\% of the original data.

cs.LG↗

arXiv2025

DUSE: A Data Expansion Framework for Low-resource Automatic Modulation Recognition based on Active Learning

Although deep neural networks have made remarkable achievements in the field of automatic modulation recognition (AMR), these models often require a large amount of labeled data for training. However, in many practical scenarios, the available target domain data is scarce and difficult to meet the needs of model training. The most direct way is to collect data manually and perform expert annotation, but the high time and labor costs are unbearable. Another common method is data augmentation. Although it can enrich training samples to a certain extent, it does not introduce new data and therefore cannot fundamentally solve the problem of data scarcity. To address these challenges, we introduce a data expansion framework called Dynamic Uncertainty-driven Sample Expansion (DUSE). Specifically, DUSE uses an uncertainty scoring function to filter out useful samples from relevant AMR datasets and employs an active learning strategy to continuously refine the scorer. Extensive experiments demonstrate that DUSE consistently outperforms 8 coreset selection baselines in both class-balance and class-imbalance settings. Besides, DUSE exhibits strong cross-architecture generalization for unseen models.

cs.LG↗

arXiv2025

FCOS: A Two-Stage Recoverable Model Pruning Framework for Automatic Modulation Recognition

With the rapid development of wireless communications and the growing complexity of digital modulation schemes, traditional manual modulation recognition methods struggle to extract reliable signal features and meet real-time requirements in modern scenarios. Recently, deep learning based Automatic Modulation Recognition (AMR) approaches have greatly improved classification accuracy. However, their large model sizes and high computational demands hinder deployment on resource-constrained devices. Model pruning provides a general approach to reduce model complexity, but existing weight, channel, and layer pruning techniques each present a trade-off between compression rate, hardware acceleration, and accuracy preservation. To this end, in this paper, we introduce FCOS, a novel Fine-to-COarse two-Stage pruning framework that combines channel-level pruning with layer-level collapse diagnosis to achieve extreme compression, high performance and efficient inference. In the first stage of FCOS, hierarchical clustering and parameter fusion are applied to channel weights to achieve channel-level pruning. Then a Layer Collapse Diagnosis (LaCD) module uses linear probing to identify layer collapse and removes the collapsed layers due to high channel compression ratio. Experiments on multiple AMR benchmarks demonstrate that FCOS outperforms existing channel and layer pruning methods. Specifically, FCOS achieves 95.51% FLOPs reduction and 95.31% parameter reduction while still maintaining performance close to the original ResNet56, with only a 0.46% drop in accuracy on Sig2019-12. Code is available at https://github.com/yaolu-zjut/FCOS.

cs.LG↗

arXiv2025

MCLRL: A Multi-Domain Contrastive Learning with Reinforcement Learning Framework for Few-Shot Modulation Recognition

With the rapid advancements in wireless communication technology, automatic modulation recognition (AMR) plays a critical role in ensuring communication security and reliability. However, numerous challenges, including higher performance demands, difficulty in data acquisition under specific scenarios, limited sample size, and low-quality labeled data, hinder its development. Few-shot learning (FSL) offers an effective solution by enabling models to achieve satisfactory performance with only a limited number of labeled samples. While most FSL techniques are applied in the field of computer vision, they are not directly applicable to wireless signal processing. This study does not propose a new FSL-specific signal model but introduces a framework called MCLRL. This framework combines multi-domain contrastive learning with reinforcement learning. Multi-domain representations of signals enhance feature richness, while integrating contrastive learning and reinforcement learning architectures enables the extraction of deep features for classification. In downstream tasks, the model achieves excellent performance using only a few samples and minimal training cycles. Experimental results show that the MCLRL framework effectively extracts key features from signals, performs well in FSL tasks, and maintains flexibility in signal model selection.

cs.LG↗

arXiv2024

Reassessing Layer Pruning in LLMs: New Insights and Methods

Although large language models (LLMs) have achieved remarkable success across various domains, their considerable scale necessitates substantial computational resources, posing significant challenges for deployment in resource-constrained environments. Layer pruning, as a simple yet effective compression method, removes layers of a model directly, reducing computational overhead. However, what are the best practices for layer pruning in LLMs? Are sophisticated layer selection metrics truly effective? Does the LoRA (Low-Rank Approximation) family, widely regarded as a leading method for pruned model fine-tuning, truly meet expectations when applied to post-pruning fine-tuning? To answer these questions, we dedicate thousands of GPU hours to benchmarking layer pruning in LLMs and gaining insights across multiple dimensions. Our results demonstrate that a simple approach, i.e., pruning the final 25\% of layers followed by fine-tuning the \texttt{lm\_head} and the remaining last three layer, yields remarkably strong performance. Following this guide, we prune Llama-3.1-8B-It and obtain a model that outperforms many popular LLMs of similar size, such as ChatGLM2-6B, Vicuna-7B-v1.5, Qwen1.5-7B and Baichuan2-7B. We release the optimal model weights on Huggingface, and the code is available on GitHub.

cs.LG↗

arXiv2024

MDM: Advancing Multi-Domain Distribution Matching for Automatic Modulation Recognition Dataset Synthesis

Recently, deep learning technology has been successfully introduced into Automatic Modulation Recognition (AMR) tasks. However, the success of deep learning is all attributed to the training on large-scale datasets. Such a large amount of data brings huge pressure on storage, transmission and model training. In order to solve the problem of large amount of data, some researchers put forward the method of data distillation, which aims to compress large training data into smaller synthetic datasets to maintain its performance. While numerous data distillation techniques have been developed within the realm of image processing, the unique characteristics of signals set them apart. Signals exhibit distinct features across various domains, necessitating specialized approaches for their analysis and processing. To this end, a novel dataset distillation method--Multi-domain Distribution Matching (MDM) is proposed. MDM employs the Discrete Fourier Transform (DFT) to translate timedomain signals into the frequency domain, and then uses a model to compute distribution matching losses between the synthetic and real datasets, considering both the time and frequency domains. Ultimately, these two losses are integrated to update the synthetic dataset. We conduct extensive experiments on three AMR datasets. Experimental results show that, compared with baseline methods, our method achieves better performance under the same compression ratio. Furthermore, we conduct crossarchitecture generalization experiments on several models, and the experimental results show that our synthetic datasets can generalize well on other unseen models.

cs.LG↗

arXiv2024

A Generic Layer Pruning Method for Signal Modulation Recognition Deep Learning Models

With the successful application of deep learning in communications systems, deep neural networks are becoming the preferred method for signal classification. Although these models yield impressive results, they often come with high computational complexity and large model sizes, which hinders their practical deployment in communication systems. To address this challenge, we propose a novel layer pruning method. Specifically, we decompose the model into several consecutive blocks, each containing consecutive layers with similar semantics. Then, we identify layers that need to be preserved within each block based on their contribution. Finally, we reassemble the pruned blocks and fine-tune the compact model. Extensive experiments on five datasets demonstrate the efficiency and effectiveness of our method over a variety of state-of-the-art baselines, including layer pruning and channel pruning methods.

cs.LG↗

arXiv2022

Mixing Signals: Data Augmentation Approach for Deep Learning Based Modulation Recognition

With the rapid development of deep learning, automatic modulation recognition (AMR), as an important task in cognitive radio, has gradually transformed from traditional feature extraction and classification to automatic classification by deep learning technology. However, deep learning models are data-driven methods, which often require a large amount of data as the training support. Data augmentation, as the strategy of expanding dataset, can improve the generalization of the deep learning models and thus improve the accuracy of the models to a certain extent. In this paper, for AMR of radio signals, we propose a data augmentation strategy based on mixing signals and consider four specific methods (Random Mixing, Maximum-Similarity-Mixing, $\theta-$Similarity Mixing and n-times Random Mixing) to achieve data augmentation. Experiments show that our proposed method can improve the classification accuracy of deep learning based AMR models in the full public dataset RML2016.10a. In particular, for the case of a single signal-to-noise ratio signal set, the classification accuracy can be significantly improved, which verifies the effectiveness of the methods.

eess.SP↗

arXiv2022

Burton-Cabrera-Frank theory for surfaces with alternating step types

Burton-Cabrera-Frank (BCF) theory has proven to be a versatile framework to relate surface morphology and dynamics during crystal growth to the underlying mechanisms of adatom diffusion and attachment at steps. For an important class of crystal surfaces, including the basal planes of hexagonal close-packed and related systems, the steps in a sequence on a vicinal surface can exhibit properties that alternate from step to step. Here we develop BCF theory for such surfaces, relating observables such as alternating terrace widths as a function of growth conditions to the kinetic coefficients for adatom attachment at steps. We include the effects of step transparency and step-step repulsion. A general solution is obtained for the dynamics of the terrace widths, assuming quasi-steady-state adatom distributions on the terraces. An explicit simplified analytical solution is obtained under widely applicable approximations. From this we obtain expressions for the full-steady-state terrace fraction as a function of growth rate. Fits of the theoretical predictions to recent experimental determinations of the steady-state and dynamics of terrace fractions on GaN (0001) surfaces during organo-metallic vapor phase epitaxy give values of the kinetic coefficients for this system. In Appendices, we also connect a model for diffusion between kinks on steps to the model for diffusion between steps on terraces, which quantitatively relates step transparency to the kinetics of atom attachment at kinks, and consider limiting cases of diffusion-limited, attachment-limited, and mixed kinetics.

cond-mat.mtrl-sci↗

arXiv2021

Deep Transfer Clustering of Radio Signals

Modulation recognition is an important task in radio signal processing. Most of the current researches focus on supervised learning. However, in many real scenarios, it is difficult and cost to obtain the labels of signals. In this letter, we turn to the more challenging problem: can we cluster the modulation types just based on a large number of unlabeled radio signals? If this problem can be solved, we then can also recognize modulation types by manually labeling a very small number of samples. To answer this problem, we propose a deep transfer clustering (DTC) model. DTC naturally integrates feature learning and deep clustering, and further adopts a transfer learning mechanism to improve the feature extraction ability of an embedded convolutional neural network (CNN) model. The experiments validate that our DTC significantly outperforms a number of baselines, achieving the state-of-the-art performance in clustering radio signals for modulation recognition.

eess.SP↗

arXiv2021

Adaptive Visibility Graph Neural Network and its Application in Modulation Classification

Our digital world is full of time series and graphs which capture the various aspects of many complex systems. Traditionally, there are respective methods in processing these two different types of data, e.g., Recurrent Neural Network (RNN) and Graph Neural Network (GNN), while in recent years, time series could be mapped to graphs by using the techniques such as Visibility Graph (VG), so that researchers can use graph algorithms to mine the knowledge in time series. Such mapping methods establish a bridge between time series and graphs, and have high potential to facilitate the analysis of various real-world time series. However, the VG method and its variants are just based on fixed rules and thus lack of flexibility, largely limiting their application in reality. In this paper, we propose an Adaptive Visibility Graph (AVG) algorithm that can adaptively map time series into graphs, based on which we further establish an end-to-end classification framework AVGNet, by utilizing GNN model DiffPool as the classifier. We then adopt AVGNet for radio signal modulation classification which is an important task in the field of wireless communication. The simulations validate that AVGNet outperforms a series of advanced deep learning methods, achieving the state-of-the-art performance in this task.

cs.LG↗

arXiv2021

CLPVG: Circular limited penetrable visibility graph as a new network model for time series

Visibility Graph (VG) transforms time series into graphs, facilitating signal processing by advanced graph data mining algorithms. In this paper, based on the classic Limited Penetrable Visibility Graph (LPVG) method, we propose a novel nonlinear mapping method named Circular Limited Penetrable Visibility Graph (CLPVG). The testing on degree distribution and clustering coefficient on the generated graphs of typical time series validates that our CLPVG is able to effectively capture the important features of time series and has better anti-noise ability than traditional LPVG. The experiments on real-world time-series datasets of radio signal and electroencephalogram (EEG) also suggest that the structural features provided by CLPVG, rather than LPVG, are more useful for time-series classification, leading to higher accuracy. And this classification performance can be further enhanced through structural feature expansion by adopting Subgraph Networks (SGN). All of these results validate the effectiveness of our CLPVG model.

cs.LG↗

arXiv2021

Crystal Truncation Rods from Miscut Surfaces with Alternating Terminations

Miscut surfaces of layered crystals can exhibit a stair-like sequence of terraces having periodic variation in their atomic structure. For hexagonal close-packed and related crystal structures with an αβαβ stacking sequence, there have been long-standing questions regarding how the differences in adatom attachment kinetics at the steps separating the terraces affect the fractional coverage of α vs. β termination during crystal growth. To demonstrate how surface X-ray scattering can help address these questions, we develop a general theory for the intensity distributions along crystal truncation rods (CTRs) for miscut surfaces with a combination of two terminations. We consider half-unit-cell-height steps, and variation of the coverages of the terraces above each step. Example calculations are presented for the GaN (0001) surface with various reconstructions. These show which CTR positions are most sensitive to the fractional coverage of the two terminations. We compare the CTR profiles for exactly oriented surfaces to those for vicinal surfaces having a small miscut angle, and investigate the circumstances under which the CTR profile for an exactly oriented surface is equal to the sum of the intensities of the corresponding family of CTRs for a miscut surface.

cond-mat.mtrl-sci↗

arXiv2021

In Situ Microbeam Surface X-ray Scattering Reveals Alternating Step Kinetics During Crystal Growth

The stacking sequence of hexagonal close-packed and related crystals typically results in steps on vicinal {0001} surfaces that have alternating A and B structures with different growth kinetics. However, because it is difficult to experimentally identify which step has the A or B structure, it has not been possible to determine which has faster adatom attachment kinetics. Here we show that in situ microbeam surface X-ray scattering can determine whether A or B steps have faster kinetics under specific growth conditions. We demonstrate this for organo-metallic vapor phase epitaxy of (0001) GaN. X-ray measurements performed during growth find that the average width of terraces above A steps increases with growth rate, indicating that attachment rate constants are higher for A steps, in contrast to most predictions. Our results have direct implications for understanding the atomic-scale mechanisms of GaN growth and can be applied to a wide variety of related crystals.

cond-mat.mtrl-sci↗

arXiv2020

Dynamics of Atomic Steps on GaN (0001) during Vapor Phase Epitaxy

Images of the morphology of GaN (0001) surfaces often show half-unit-cell-height steps separating a sequence of terraces having alternating large and small widths. This can be explained by the $αβαβ$ stacking sequence of the wurtzite crystal structure, which results in steps with alternating $A$ and $B$ edge structures for the lowest energy step azimuths, i.e. steps normal to $[0 1 \bar{1} 0]$ type directions. Predicted differences in the adatom attachment kinetics at $A$ and $B$ steps would lead to alternating $α$ and $β$ terrace widths. However, because of the difficulty of experimentally identifying which step is $A$ or $B$, it has not been possible to determine the absolute difference in their behavior, e.g. which step has higher adatom attachment rate constants. Here we show that surface X-ray scattering can measure the fraction of $α$ and $β$ terraces, and thus unambiguously differentiate the growth dynamics of $A$ and $B$ steps. We first present calculations of the intensity profiles of GaN crystal truncation rods (CTRs) that demonstrate a marked dependence on the $α$ terrace fraction $f_α$. We then present surface X-ray scattering measurements performed \textit{in situ} during homoepitaxial growth on (0001) GaN by vapor phase epitaxy. By analyzing the shapes of the $(1 0 \bar{1} L)$ and $(0 1 \bar{1} L)$ CTRs, we determine that the steady-state $f_α$ increases at higher growth rate, indicating that attachment rate constants are higher at $A$ steps than at $B$ steps. We also observe the dynamics of $f_α$ after growth conditions are changed. The results are analyzed using a Burton-Cabrera-Frank model for a surface with alternating step types, to extract values for the kinetic parameters of $A$ and $B$ steps. These are compared with predictions for GaN (0001).

cond-mat.mtrl-sci↗

arXiv2018

Persistence of Island Arrangements During Layer-by-Layer Growth Revealed Using Coherent X-rays

Understanding surface dynamics during epitaxial film growth is key to growing high quality materials with controllable properties. X-ray photon correlation spectroscopy (XPCS) using coherent x-rays opens new opportunities for in situ observation of atomic-scale fluctuation dynamics during crystal growth. Here, we present the first XPCS measurements of 2D island dynamics during homoepitaxial growth in the layer-by-layer mode. Analysis of the results using two-time correlations reveals a new phenomenon - a memory effect in island nucleation sites on successive crystal layers. Simulations indicate that this persistence in the island arrangements arises from communication between islands on different layers via adatoms. With the worldwide advent of new coherent x-ray sources, the XPCS methods pioneered here will be widely applicable to atomic-scale processes on surfaces.

cond-mat.mtrl-sci↗

arXiv2018

Spontaneous Repairing Liquid Metal/Si Nanocomposite as a Smart Conductive-Additive-Free Anode for Lithium-ion Battery

Silicon is a promising candidate for negative electrodes due to its high theoretical specific capacity (~3579 mAh g-1) and low lithiation potential (~0.40 V vs Li). However, its practical applications in battery have been inhibited by the large volume change (~400%) induced by Li+-insertion into Si lattices. Here, we attempt to resolve this issue at a fundamental level, and report for the first time a novel liquid metal (LM)-mediated spontaneous repairing conductive-additive-free Si anode for Li-ion battery. The fluidity of LM ensures the eternal contact between Si and the conducting-network during its repeated electrochemical reactions. The as-prepared nano-composite of LM/Si leads to superior performances as characterized by high capacity utilization (2300 mAh g-1 at 500 mA g-1), long-term stability (968 mAh g-1 after 1500 charge-discharge cycles at 8 A g-1 with 81.3% retention), high rate capability (360 mAh g-1 at 20 A g-1, equivalence of 55 C, or full charge/discharge in 65 seconds), and, in particular, an extra-ordinarily high initial coulombic efficiency (95.92%), which is not only the highest reported for Si to the best of our knowledge, but also higher than the mature graphitic carbon anodes. The unique approach described in this work not only resolves the basic stress challenges faced by the promising but often problematic alloy-type materials; in broader context it also provides a universal inspiration to all electrode materials whose electric properties suffer from extreme mechanic upheavals induced by the electrochemical strains during the cell reactions.

physics.app-ph↗