A5025147440- Wireless Signal Modulation Classification
- Radar Systems and Signal Processing
- Advanced SAR Imaging Techniques
- Metamaterials and Metasurfaces Applications
- Machine Learning in Bioinformatics
- Radio Wave Propagation Studies
- Indoor and Outdoor Localization Technologies
- Ultrasonics and Acoustic Wave Propagation
Xidian University
2022-2023
In this work, the Transformer Network (TRN) is applied to automatic modulation classification (AMC) problem for first time. Different from other deep networks, TRN can incorporate global information of each sample sequence and exploit that semantically relevant classification. order illustrate performance proposed model, it compared with four models two traditional methods. Simulation results show one has a higher accuracy especially at low signal noise ratios (SNRs), number training...
The existing research on deep learning for radar signal intra–pulse modulation classification is mainly based supervised leaning techniques, which performance relies a large number of labeled samples. To overcome this limitation, self–supervised framework, contrastive (CL), combined with the convolutional neural network (CNN) and focal loss function proposed, called CL––CNN. A two–stage training strategy adopted by CL–CNN. In first stage, model pretrained using abundant unlabeled...
Radar intrapulse signal modulation classification is an important work for the electronic countermeasure and there are mainly two categories of algorithms. The deep learning-based algorithms usually outperform traditional feature extraction-based ones, but they may rely on massive labeled samples training, which limits their practical applications. To solve this problem, SS-LWCNN model combines semisupervised learning (SI-SL) with virtual adversarial training (VAT) light weight technology...
In this work, the omni-dimensional dynamic convolution (ODConv) layer based network (OD-CNN) with focal loss function is applied to radar intra-pulse signal modulation classification, which greatly improves classification accuracy. Compared layer, ODConv employs a novel multi-dimensional attention mechanism learn four types of attentions along dimensions kernel space in parallel manner, further feature mining ability model. order illustrate superior proposed model, it compared other three...
In this paper, the two-dimensional (2-D) programmable sparse metasurface (PSM) is constructed, and direction of arrival (DOA) estimation algorithm based on it proposed. The rows are located with spacings according to a sequence, then equivalent covariance matrix an uniform more virtual patches which improves estimated number 2-D DOAs DOA accuracy. simulation results show that performance proposed PSM almost equal greater patches.