A comprehensive and well-structured review on the application of deep learning (DL) based algorithms, such as convolutional neural networks (CNN) and long-short term memory (LSTM), in radar signal processing is given. The following DL application areas are covered: i) radar waveform and antenna array design; ii) passive or low probability of interception (LPI) radar waveform recognition; iii) automatic target recognition (ATR) based on high range resolution profiles (HRRPs), Doppler signatures, and synthetic aperture radar (SAR) images; and iv) radar jamming/clutter recognition and suppression. Although DL is unanimously praised as the ultimate solution to many bottleneck problems in most of existing works on similar topics, both the positive and the negative sides of stories about DL are checked in this work. Specifically, two limiting factors of the real-life performance of deep neural networks (DNNs), limited training samples and adversarial examples, are thoroughly examined. By investigating the relationship between the DL-based algorithms proposed in various papers and linking them together to form a full picture, this work serves as a valuable source for researchers who are seeking potential research opportunities in this promising research field.

The key to the feasibility of spectrum sharing between radar and wireless systems is effective mainlobe interference mitigation processing by the radar. Traditional array processing is capable to cancel sidelobe interferences only. This letter presents an innovative beamforming approach, the first of its kind, to eliminate wireless interference in both mainlobe and sidelobe directions based on a coherent phase-coding multiple-input multiple-output (MIMO) radar platform. The theoretical proof and conditions of simultaneous mainlobe interference cancellation and target detection using coherent MIMO radar are mathematically derived and numerically verified. The simulation results demonstrate the radar can effectively eliminate wireless interferences from base stations and mobile handsets during spectrum-sharing.

Mass production of high-quality synthetic SAR training imagery is essential for boosting the performance of deep-learning (DL)-based SAR automatic target recognition (ATR) algorithms in an open-world environment. To address this problem, we exploit both the widely used Moving and Stationary Target Acquisition and Recognition (MSTAR) SAR dataset and the Synthetic and Measured Paired Labeled Experiment (SAMPLE) dataset, which consists of selected samples from the MSTAR dataset and their computer-generated synthetic counterparts. A series of data augmentation experiments are carried out. First, the sparsity of the scattering centers of the targets is exploited for new target pose synthesis. Additionally, training data with various clutter backgrounds are synthesized via clutter transfer, so that the neural networks are better prepared to cope with background changes in the test samples. To effectively augment the synthetic SAR imagery in the SAMPLE dataset, a novel contrast-based data augmentation technique is proposed. To improve the robustness of neural networks against out-of-distribution (OOD) samples, the SAR images of ground military vehicles collected by the self-developed MiniSAR system are used as the training data for the adversarial outlier exposure procedure. Simulation results show that the proposed data augmentation methods are effective in improving both the target classification accuracy and the OOD detection performance. The purpose of this work is to establish the foundation for large-scale, open-field implementation of DL-based SAR-ATR systems, which is not only of great value in the sense of theoretical research, but is also potentially meaningful in the aspect of military application.

Coherent multiple-input multiple-output (MIMO) radar is capable of forming a specific transmit beam pattern favorable for radar operation, but the transmitted waveforms from different antenna elements need to be orthogonal for coherent beamforming. In this work, an innovative approach is developed for MIMO radar waveform design in space-time domain, rendering MIMO radar able to satisfy transmit beamforming constraint in space-domain as well as the waveform orthogonality requirement in time domain. The approach is validated through simulations.

For coherent MIMO radar the optimal target signal processing can be achieved for any transmitted waveforms or radiation beam pattern, making transmit beamforming through waveform design possible without degrading target detection performance. In this work, an innovative waveform optimization approach termed phase-only variable metric method (POVMM) is proposed for coherent MIMO radar waveform design to form a desired transmit beam pattern such as one with radiation nulls in certain directions. The waveform design is carried out by minimizing the radiation powers of the MIMO radar antenna in the selected directions with optimization variables constrained to the waveform phases only. The gradient function of the cost function with regard to waveform phases is analytically derived for the optimization and the POVMM is developed based on the variable metric methods with a flexible search step sizing strategy for improving optimization efficiency. The proposed approach is validated with various designs and simulations.

The netted radar systems are known for being able to provide higher target detection probabilities, more accurate measurements, and better multi-target tracking performance than the monostatic/bistatic radar systems by exploiting information fusion processing. Simple radar networks made of multiple independent monostatic radars with decentralized processing have been around for several decades. In contrast, sophisticated radar networks with centralized data fusion only emerged during the last few years thanks to the development in high-speed digital processing and precise time/phase synchronization. Although the advanced radar networks offer the potential ability to detect stealth targets and low probability of intercept radar signal, there are many technical challenges to be solved before they could evolve from concepts in the research paper to implementable systems in the real-world. By presenting the various target detection and parameter estimation approaches adopted by the distributed MIMO radar, the passive radar network, the hybrid active-passive radar network, and the MIMO synthetic aperture radar (SAR), we lay out the foundation for the future research to be conducted to solve the key issues limiting the practical performance of these advanced radar networks.