Deep learning (DL) is a new machine (ML) methodology that has found successful implementations in many application domains. However, its usage communications systems not been well explored. This paper investigates the use of DL modulation classification, which major task systems. The relies on massive amount data and, for research and applications, this can be easily available Furthermore, unlike ML, advantage requiring manual feature selections, significantly reduces complexity...

A computer-aided diagnosis (CAD) system based on mammograms enables early breast cancer detection, diagnosis, and treatment. However, the accuracy of existing CAD systems remains unsatisfactory. This paper explores a method feature fusion with convolutional neural network (CNN) deep features. First, we propose mass detection CNN features unsupervised extreme learning machine (ELM) clustering. Second, build set fusing features, morphological texture density Third, an ELM classifier is...

Deep learning (DL) is a powerful classification technique that has great success in many application domains. However, its usage communication systems not been well explored. In this paper, we address the issue of using DL systems, especially for modulation classification. Convolutional neural network (CNN) utilized to complete task. We convert raw modulated signals into images have grid-like topology and feed them CNN training. Two existing approaches, including cumulant support vector...

Channel estimation is a challenging task, especially in massive multiple-input multiple-output (MIMO) system with one-bit analog-to-digital converters (ADC). Traditional deep learning (DL) methods, that learn the mapping from inputs to real channels, have significant difficulties estimating accurate channels because their loss functions are not well designed and investigated. In this paper, conditional generative adversarial networks (cGAN) developed predict more realistic by adversarially...

We propose Denoising Masked Autoencoder (Deno-MAE), a novel multimodal autoencoder framework for denoising modulation signals during pretraining. DenoMAE extends the concept of masked autoencoders by incorporating multiple input modalities, including noise as an explicit modality, to enhance cross-modal learning and improve performance. The network is pre-trained using unlabeled noisy constellation diagrams, effectively reconstruct their equivalent noiseless diagrams. Deno-MAE achieves...

Manual lung ultrasound (LUS) scoring is influenced by clinicians' subjective interpretation, leading to potential inconsistencies and misdiagnoses due varying levels of experience. To improve monitoring pulmonary ventilation support early diagnosis, we propose an automated LUS network based on 8-point scale, named the detailed-global fusion residual (DGF-ResNet). This combines local global features using hybrid feature Block, which includes detail extraction (DFE) (GFE) Modules. The DFE...

// Ying Liao 1, * , Hua-Xia Wang 3, Xiang Mao 1 Hongjie Fang 3 Huang Yanrong Li Yingjie Sun Chun Meng Lei Tan Cuiping Song Xusheng Qiu and Chan Ding 2 Department of Avian Diseases, Shanghai Veterinary Research Institute, Chinese Academy Agricultural Sciences, 200241, P. R. China Jiangsu Co-innovation Center for Prevention Control Important Animal Infectious Diseases Zoonoses, Yangzhou 225009, College Medicine, Nanjing University, 210095, These authors have contributed equally as first author...

In cognitive radio (CR) networks, cooperative spectrum sensing is utilized to improve the performance avoid potential interference primary users (PUs) and increase access opportunities for secondary (SUs). A process divided into three phases: individual sensing/detection, reporting/fusion, data transmission. reporting phase, one or more channels are needed transmit results a fusion center (FC), global determined at FC. The number of required depends on sensors SUs, which relates channel...

Asp-Glu-Ala-Asp (DEAD)-box RNA helicase 3 (DDX3), an ATP-dependent helicase, is associated with splicing, mRNA export, transcription, translation, and decay. Recent studies revealed that DDX3 participates in innate immune response during virus infection by interacting TBK1 regulating the production of IFN-β. In our studies, we demonstrated regulated NF-κB signal pathway. We found knockdown reduced phosphorylation p65 IKK-β ultimately attenuated inflammatory cytokines induced poly(I:C) or...

Viral infections result in cellular stress responses, which can trigger protein translation shutoff via phosphorylation of eukaryotic initiation factor 2 alpha (eIF2α). Newcastle disease virus (NDV) causes severe poultry and selectively kills human tumour cells. In this report, we determined that infection HeLa cervical cancer cells DF-1 chicken fibroblast with NDV maintained at early times, 0–12 h post-infection (p.i.), gradually inhibited global late 12–24 p.i. Mechanistic studies showed...

Deep neural networks have been shown to perform well in many classical machine learning problems, especially image classification tasks. However, researchers found that can be easily fooled, and they are surprisingly sensitive small perturbations imperceptible humans. Carefully crafted input images (adversarial examples) force a well-trained network provide arbitrary outputs. Including adversarial examples during training is popular defense mechanism against attacks. In this paper we propose...

Cognitive radio is an effective technology to improve spectrum utilization through sharing among primary users (PUs) and secondary (SUs). Developing accurate efficient methods determine the PU coverage crucial in order avoid potential interference from SU PU. In this paper, we propose use estimated location of each SU, which obtained using self-organizing maps algorithm, boundary coverage. Furthermore, detection issue considering presence malicious SUs also investigated. Support vector...

Differences in acquisition time, light conditions, and viewing angle create significant differences among the airborne remote sensing images from Unmanned Aerial Vehicles (UAVs). Real-time scene matching navigation applications based on fixed reference maps are error-prone have poor robustness. This paper presents a novel shadow-based method for localization of low-altitude flight UAVs. A shadow map is generated an accurate (0.5 m spatial resolution) Digital Surface Model (DSM) with known...

A coordinated jamming and communications technique for degrading enemy user transmission performance in a multicarrier code-division multiple-access system is introduced. linear minimum mean square error (MMSE) algorithm with matrix estimation single-tone presented. Furthermore, blind MMSE multiuser detection multitone investigated. Simulation results show that friendly users are able to achieve reliable communications, while the of significantly degraded.

In a massive multiple-input-multiple-output (MIMO) uplink system, the pilot sequence reuse in neighboring cells causes contamination, causing decoding performance to degrade significantly. this paper, blind method based on independent component analysis (ICA) is proposed without using sequences. The uses ICA separate received signals (from in-cell and cells) estimate channels. energy levels of estimated channels are used differentiate an signal from cell signals. analytical results derived....

A massive multiple-input multiple-output (MIMO) system, which utilizes a large number of antennas at base stations to communicate with multiple user terminals each single antenna, is one the most promising techniques for future wireless communications systems. successful MIMO implementation relies on accurate channel estimation, typically performed through pilot sequences. However, estimation performance or limited by contamination due unavoidable reuse sequences from in neighboring cells....

Due to its portability, convenience, and low cost, incompletely closed near-infrared (ICNIR) imaging equipment (mixed light reflection imaging) is used for ultra thin sensor modules have good application prospects. However, with structure also brings some problems. Some finger vein images are not clear there sparse or even missing veins, which results in poor recognition performance. For these quality ICNIR images, however, additional fingerprint information the image. The analysis of...