Owing to the large degrees of freedom and reduced mutual coupling by generating difference coarrays, nonuniform linear arrays have aroused great interest in direction arrival estimation. Previous works shown some improved sparse arrays, while few find common features hidden within these structures. In this paper, we define a generic-coarray concept reveal impacts variable ranges element spacing on uniform (uDOFs), which sufficient condition for connected coarrays is derived. We then propose...

Recently, deep unfolded networks have been widely utilized in direction of arrival (DOA) estimation due to the reduced computational complexity and improved accuracy. However, few consider nested array for off-grid DOA estimation, where estimated DOAs are not on prespecified grids. In this article, we propose a FOCal underdetermined system solver network alternating method multiplies address problem, which respectively aim improve accuracy further reduce complexity. We first apply...

Deep learning methods have played a more important role in hyperspectral image classification. However, general deep mainly take advantage of the samplewise information to formulate training loss while ignoring intrinsic data structure each class. Due high spectral dimension and great redundancy between different channels image, these former losses usually cannot work so well for representation image. To tackle this problem, develops novel manifold embedding method (DMEM) First, class is...

In this paper, the convolution neural networks (CNN) are developed for classification and localization of mixed near-field far-field sources by using geometry symmetric nested array. We first transform received data into frequency domain. Then, we preprocess phase difference matrix to decouple sources. Considering that counter-diagonal elements in only contain direction arrival (DOA) parameter each source, utilize upper right as input CNN estimate DOA source. order avoid influence noise,...

Recently, deep unfolding networks have been widely used in direction of arrival (DOA) estimation because their improved accuracy and reduced computational cost. However, few considered the existence a nested array (NA) with off-grid DOA estimation. In this study, we present sparse Bayesian learning (DSBL) network to solve problem. We first establish signal model for NA. Then, transform output into real domain neural networks. Finally, construct train DSBL determine on-grid spatial spectrum...

Recently, deep unfolding networks with interpretable parameters have been widely utilized in direction of arrival (DOA) estimation due to the faster convergence speed and better generalization ability. However, few consider nested array for gridless DOA estimation. In this letter, we propose a alternating projection network address problem. We first convert covariance matrix into measurement vector form atomic norm, which can reduce dimension during projection. then train proposed...

Abstract Deep Neural Networks have been successfully applied in hyperspectral image classification. However, most of prior works adopt general deep architectures while ignore the intrinsic structure image, such as physical noise generation. This would make these models unable to generate discriminative features and provide impressive classification performance. To leverage information, this work develops a novel learning framework with inclined module denoise for First, spectral signature is...

Deep Neural Networks (DNNs), from AlexNet to ResNet ChatGPT, have made revolutionary progress in recent years, and are widely used various fields. The high performance of DNNs requires a huge amount high-quality data, expensive computing hardware, excellent DNN architectures that costly obtain. Therefore, trained becoming valuable assets must be considered the Intellectual Property (IP) legitimate owner who created them, order protect models illegal reproduction, stealing, redistribution, or...

Coprime arrays (CPAs) have been found to be an effective configuration for adaptive beamforming (ABF). However, most ABF methods CPAs are embedded with the spatial spectrum estimation technique, and therefore they usually deteriorate in case of low signal-to-noise ratio (SNR), especially without prior information interferences. To address this issue, a robust method via atomic-norm-based sparse recovery is proposed. This begins initialised virtual array interpolation avoid aperture loss...

Recently, passive bistatic radar (PBR) has become an emerging technology because it allows target detection and localization with advantages of low cost, covert detection, vulnerability to electronic jamming high valuable counter stealth technology. However, these researchers mainly focus on the illuminators opportunity such as FM radio, television broadcast, mobile network basestation, wireless fidelity, long term evolution, global navigation satellite, etc. Different from aforementioned...

In the case of multipath propagation for passive bistatic radar (PBR) using uncooperative frequency agile-phased array as an illuminator, a new direction-finding method is proposed to deal with scenario where coherent and uncorrelated signals are closely spaced or in same direction. Firstly, spatial difference technique used eliminate signals. Then, order avoid cross-terms effect improve resolution signal, iterative adaptive (IAA) adopted rearranged matrix. Finally, direction arrival (DOA)...

Direction-of-arrival (DOA) estimation in multipath environment is an important issue for passive bistatic radar (PBR) using frequency agile phased array VHF as illuminator of opportunity. Under such scenario, the main focus this paper to cope with closely spaced uncorrelated and coherent signals low signal-to-noise ratio limited snapshots. Making full use characteristics moduli eigenvalues, DOAs are firstly estimated. Afterwards, their contributions eliminated by means spatial difference...

Passive bistatic radars (PBR) using illuminators of opportunity have attracted much attention over the last years. However, most PBR systems been rather experimental set-ups tailored to a single frequency band or implemented as laboratory test devices. In case exploiting transmitters for sources illumination, there is problem that radar transmitting signal parameters are unknown, and processing cannot be conducted accurately. This paper presents study demonstrate feasibility Digital Array...

In case of passive bistatic radar (PBR) using a phased array (PAR) as the non-cooperative illuminator, there are problems that signal parameters transmitter unknown and coherence is damaged. Aiming at issues, this article presents target detection method for PBR PAR illuminator. Firstly, uses direct wave to estimate parameters, then pulse compresses scattered wave. Secondly, compressed aligned by fast-time dimension so Range-Time (R-T) map can be obtained. Finally, detected according...

Passive bistatic radar (PBR) has attracted widespread attention for its capabilities in dealing with the threat of electronic countermeasure, stealth technology, and antiradiation missile. However, passive detection methods are limited by unknown characteristics uncooperative illuminators, conventional signal processing algorithms cannot be conducted accurately, especially when carrier frequency transmitting is agile signal-to-noise ratio (SNR) scattered wave target low. To address above...

In irregular pulse repetition interval (PRI) radar, successive pulses each with different PRIs are used as the transmission waveform. After analyzing signal model of PRI we propose a coherent integration method based on Radon-iterative adaptive approach (Radon-IAA) to deal problems range cell migration (RCM) and phase fluctuations among introduced by PRI. our method, RCM is compensated searching through motion parameters, can be resolved an IAA-based spectral analysis method. The...

Owing to the large degrees of freedom and reduced mutual coupling by producing difference coarrays, nonuniform linear arrays have aroused great interest in direction arrival (DOA) estimation. Previous works presented some new sparse arrays, such as thinned coprime array. In this paper, we propose a generalized array introducing flexible inter-element spacings, where conventional one can be seen special case. We derive closedform expression for range consecutive lags, written functions...

High Resolution Range Profiles (HRRP) have become a key area of focus in the domain Radar Automatic Target Recognition (RATR). Despite success data-driven neural network-based HRRP recognition, challenges such as insufficient training samples persist its real-world application. This letter introduces HRRPGraphNet, novel Graph Neural Network (GNN) model designed specifically for target recognition that leverages new insights to address these challenges. A pivotal innovation is transformation...

This paper introduces an innovative deep learning-based method for end-to-end target radial length estimation from HRRP (High Resolution Range Profile) sequences. Firstly, the sequences are normalized and transformed into GAF (Gram Angular Field) images to effectively capture utilize temporal information. Subsequently, these serve as input a pretrained ResNet-101 model, which is then fine-tuned estimation. The simulation results show that compared traditional threshold simple networks e.g....