Search

Classifiers based on machine learning are usually trained to distinguish between several known classes. For an electronic intelligence application, however, it is of great importance to recognize if an intercepted signal belongs to an unknown radar emitter. In the machine learning literature, this task is called open-set recognition. This article investigates six approaches in several configurations to recognize unknown emitters. It is based on a hierarchical emission model that understands emissions as a language with an inherent hierarchical structure. We consider two general approaches, which are the ""memoryless"" Markov chain and the Long Short-Term Memory recurrent neural network, which is especially designed to ""remember"" the past. The performance is demonstrated with two evaluation metrics in ten scenarios that contain different combinations of known and unknown emitters. An evaluation with corrupted data provides an estimate on the methods' accuracies under challenging conditions. The results show that unknown emitters that do not use known waveforms are reliably recognized even with corrupted data, while unknown emitters that are more similar to known ones are harder to detect.

Radar specific emitter identification (SEI) is the process of uniquely identifying an individual emitter from the same class of radars by their individual properties that arise from hardware imperfections. However, it is challenging or perhaps impossible to generate globally unique emitter identifiers by using SEI techniques alone, due to the increasing number of radars and the subtle differences in their signal properties. We therefore introduce a multitier radar emitter identification concept that includes radar function identification, model identification, and SEI. The combination of function identifiers, model identifiers, and specific emitter identifiers could generate globally unique radar emitter identifiers. In this article, we propose to use the radio frequency (RF) features extracted from time domain transient signals for radar model identification. The RF features include the duration, maximum derivative, skewness, kurtosis, mean, variance, fractal dimension, Shannon entropy, and polynomial coefficients of the normalized energy trajectory of a transient signal, as well as the area under the trajectory curve. We propose three RF fingerprints for radar model identification, each consisting of a predetermined subset of the features. The performance of the RF fingerprints was evaluated by using five classification algorithms with two radar datasets. Our results show attractive performance with respect to hetero-model radar identification. In particular, for the hetero-model radar dataset, the pair of the all-features (AF) fingerprint and gradient boosting algorithm achieved 91.9, 90.25, and 91.1% classification accuracy for three, four, and five emitters, respectively. On this basis, we conclude that the proposed AF fingerprint could be applied directly to radar model identification by training the gradient boosting algorithm using a dataset with many radars per model.

This article concerns the issue of how to combine the multidomainradar information, including range–Doppler, time–Doppler, and time–range, for human activity recognition. Specifically, to fully make use of radar information, instead of using a single-domain spectrum as inputs, a novel hybrid neural network model is developed for exploring multidomain fusion of radar information. In doing this, three kinds of 2-D domain spectra are used in a fashion of supplementing each other with a hybrid framework that combines three models: 1-D convolution neural network, recurrent neural network, and 2-D convolution network. It is advantageous to use such a hybrid model to capture much rich features through multidomain feature fusion, so as to improve the accuracy of human activity recognition effectively. Experimental results validate the proposed method.

In this article, we review and compare different methods for the localization of an illuminator of opportunity for passive radar systems employing cooperative targets. For assessment and comparison, we use numerical simulations and real-world radar measurements. To achieve reliable operation under real-world conditions, a correct and robust association of the bistatic radar target observations with ground truth information is necessary. We introduce a novel solution to this issue with an efficient preprocessing algorithm based on a planar approximation.

Phased array radar systems are indispensable in many applications requiring robust sensing of the environment. To achieve sensitive target detection and accurate direction-of-arrival (DOA) estimation, a high number of receiving antenna elements is needed. The high dimension of the element level data inevitably leads to a large computational burden for the digital signal processing. This problem can be overcome by transforming the element level data into a lower dimensional beamspace. In this paper, we present a novel parameter-controlled design method to construct this transformation. If the dimension reduction is not too drastic, it jointly achieves optimal detection and DOA estimation performance. Otherwise, it meets pre-defined performance criteria by exploiting an acceptable trade-off between detection and DOA estimation performance. We propose a general design tool, which is not limited to a specific array configuration. The design tool comprises a pre-calculated set of plots, providing the radar designer an overview of possible performance for a given scenario. We describe a straightforward method to construct the corresponding transformation. Numerical studies highlight the superiority of the proposed design method.

The detection and tracking performance of passive radar systems can be enhanced by exploiting multiple emitters and one receiver site by allowing for multiple bistatic geometries. This article investigates the measurement accuracy of passive radar systems that employ the advanced television system committee (ATSC 1.0) signal as a signal of opportunity. The monostatic and bistatic modified Cramer–Rao lower bound for the range and velocity measurements are derived as a quantitative way to evaluate the performance variation when exploiting multiple transmitters. The effect of 3D target position on the signal-to-noise ratio, with respect to stationary transmitters and receiver, is highlighted. Furthermore, signal-to-interference-and-noise ratio modeling is included by considering the direct signal interference suppression performance for a set of transmitters. An example in Columbus, Ohio, for a single receiver and multiple digital television transmitters that are spatially distributed is presented to find the optimal emitter that yields the best range and velocity measurement in the analyzed area. Target-tracking experiments show an agreement between the optimal emitter map and the tracking quality for the set of transmitters.

Accurate and stable time and frequency transfer (TFT) is typically required in distributed sensing. While this has historically been a limiting factor in the design of netted radars, the rise of modern TFT techniques has paved the way to nanosecond level time accuracy, absolute phase synchronization, and sub-Hertz frequency accuracy. In particular, netted systems composed of mobile nodes have benefited immensely from TFT based on global navigation satellite systems (GNSS). However, GNSS systems are liable to denial of service interruption. This article demonstrates the use of a fiber-optic White Rabbit (WR) network in an operational multistatic radar—NeXtRAD, and directly compares the results to one-way multichannel GPS disciplined oscillators acting under common-view TFT.

Sensor registration is an important problem that must be considered when attempting to perform any kind of data fusion in multimodal, multisensor target tracking. In this multiple target tracking (MTT) application, any inaccuracies in the registration can lead to false tracks being created, and tracks of true targets being stopped prematurely. This article introduces a method for simultaneously tracking multiple targets in a surveillance region and estimating appropriate sensor registration parameters so that sensor fusion can be performed accurately. The proposed method is based around particle belief propagation (BP), a recent but highly efficient framework for tracking multiple targets. The proposed method also uses a hierarchical model which allows for multiple processes to be linked and interact with one another. We present a comprehensive set of simulations and results using differing, asynchronous sensor setups, and compare with a random finite set (RFS) approach, namely the sequential Monte Carlo (SMC)-probability hypothesis density (PHD) filter. The results show the proposed method is 17% more accurate than the RFS approach on average.

Fast Fourier transform (FFT) is one of the fundamental signal processing algorithms widely used in radar applications. The Radon–Fourier transform (RFT) can be seen as an FFT generalization that can overcome some of its limitations. This work derives three spectral RFT (SRFT) based approaches to address major challenges of the multiple-input multiple-output automotive radars. First, two SRFT-based approaches are derived to increase maximal target detection range by mitigation of target migration in range and direction of arrival, jointly, and by multidwell integration processing, which increases the radar coherent integration time without compromising its detection update rate. Next, SRFT-based approach is proposed to address the cluster-to-track association problem that arises in multiple distributed target tracking scenarios that characterize automotive radar operation in dense urban environments.

A fast and excellent unimodular waveform with good autocorrelation and cross-correlation design method for the multiple-input multiple-output radar is devised. Unlike the existing methods that only optimize partial metrics or only optimize the short sequence in acceptable time, we propose a comprehensive waveform design method to minimize the weighted sum of almost entirely metrics under the constant modulus constraint. Then, a deep learning framework, named as the comprehensive optimization network, is derived to handle the problem. Numerical results show that the proposed method has superior performance and acceptable optimization time compared with the existing methods.

Radar mounted on a moving vehicle returns time-varying detections corresponding to unstable scattering points, unlike optical sensors, which produce relatively stable detections. We present an efficient extended Kalman filter-based simultaneous localization and mapping (EKF-SLAM) algorithm for radar, utilizing new techniques of clustering and sifting the time-varying detections. Velocity bias and yaw rate bias, which are inherent in any odometer are also estimated and compensated using the same EKF. For theoretical performance evaluation, the posterior Cramer–Rao bound (PCRB) for SLAM with odometer bias estimation is derived and compared with the root-mean-squared errors (RMSEs), with and without bias estimation. The simulation results show that the RMSE for SLAM with bias estimation is the closest to the PCRB. The proposed algorithm is also verified experimentally with field data. The comparison with the ground truth trajectory obtained from a differential global positioning system shows that the proposed algorithm yields an accurate trajectory estimate, even under large odometer bias. Also, the real-time capability of the proposed algorithm is verified.

How to utilize the limited antennas for tracking multiple targets plays a critical role in the collocated multiple-input multiple-output radar. A dynamic antenna selection strategy is proposed to address this problem. The basis of our strategy is to achieve the optimal antenna selection under the constraint of limited active antennas using the feedback information in the tracking recursion cycle, improving the worst case of estimate accuracy among multiple targets. Since the posterior Cramer–Rao lower bound quantifies the target tracking performance, it is derived and utilized as the optimization criterion. We then propose an efficient algorithm which integrates the convex relaxation technique with the local search to solve the problem. Simulation results show its superior performance compared with the random antenna selection strategy and the heuristic search algorithm. Moreover, the proposed strategy can provide the performance close to the exhaustive search method while maintaining reasonable runtime.

To perform 3-D inverse synthetic aperture radar (ISAR) imaging through interferometric processing, all ISAR images should be correctly registered such that the same scatterer appears at the same position along both the range and Doppler directions. However, this condition generally does not hold in most situations because different radar view angles introduce additional angular motion errors that must be compensated. Hence, we herein propose a new ISAR registration method based on joint-channel phase difference (JC-PD) operations. From the result of the JC-PD function, JC-PD profiles are obtained, in which the locations and phases of the maximum peak contain constant and time-varying angular motions, respectively. Using our proposed method, correct ISAR registrations can be achieved in both the range and Doppler directions. In particular, compared to conventional Doppler registration methods, time-varying angular motions are more accurately compensated without the problem of cross-term phase interference. Through simulations and experiments, we verified that the proposed method yielded accurate ISAR registration results; hence, an excellent 3-D target reconstruction through interferometric ISAR imaging was demonstrated.

Analyzing radar signals is an important task in operating electronic support measure systems. The received signals in the real electromagnetic environment often originate from multiple emitters and must be separated for further processing. Pulses from important target emitters with known parameters should be picked out first. To solve the problem, time-of-arrival (TOA) deinterleaving may be performed to extract signals from a certain emitter by learning the pulse repetition interval (PRI) modulation that makes up the signal. However, conventional deinterleaving methods only work with simple PRI modulations; their performance degrades in noisy environments. A novel approach based on denoising autoencoders for TOA deinterleaving was developed in this article. The inner patterns of pulse-of-interest sequences were learned by the proposed denoising autoencoders to generate output sequences from well-trained autoencoders. Simulation results show that the proposed method outperforms conventional methods, especially in environments with high lost and spurious pulse ratios.

Accurate orbit prediction (OP) of space debris is vital in space situation awareness (SSA) related tasks, such as space collision warnings. However, owing to the sparse and low precision observations, unknown geometrical and physical features of debris, and effects of incomplete force models, OP based on the orbital mechanics theory or physics-based OP of space debris suffers from rapid error growth over a long duration, limiting the period of validity of debris OP for precise space applications. Considering that the tracking arcs of a debris object over a single station often share a similar temporal and spatial distribution in the inertial space, the resultant OP errors possibly have a coherent relationship with the temporal and spatial distribution of tracking arcs. This article proposes a machine learning (ML)-based approach to model the underlying pattern of debris OP errors from historical observations and apply it to modify the future physics-based OP results. The approach includes three steps: constructing a historical OP error set, training an ML model to fit the historical OP error set, and correcting the future physics-based OP with ML-predicted orbital errors. The ensemble learning algorithm of boosting tree is studied as the primary ML method for the error modeling and predicting process. Experiments with three low-Earth-orbit objects, tracked by a single radar station, demonstrate that the trained ML models can capture more than 80% of the underlying pattern of the historical OP errors. More importantly, the errors of physics-based OP over the future seven days reduce from thousands of meters to hundreds or even tens of meters through the error correction with the learned error pattern, achieving at least 50% accuracy improvement. Such dramatic OP improvements show the promising potential of ML for enhanced SSA capability.

An adaptive ranging technique for maintaining high-accuracy ranging between nodes in coherent distributed antenna arrays is presented. Coherent distributed antenna arrays are networks of wireless systems coordinated coherently at the level of the wavelength of the wireless signal. Enabling coherent operation between separate mobile nodes for active and passive microwave remote sensing requires accurate knowledge of the relative positions of the nodes in the array. In this article, a novel adaptive ranging technique based on the near-optimal waveform for high-accuracy ranging, a two-tone waveform, is designed and demonstrated in software-defined radio platforms representing array nodes. Ranging accuracy is dependent on both signal-to-noise ratio and the separation of the two tones in the waveform; however, in realistic environments, factors such as attenuation or antenna misalignment are not easily predicted, which can lead to degradation of the ranging measurement. Selecting one appropriate waveform for range measurements is, thus, not feasible unless the bandwidth assigned to it is always higher than required for the needed range accuracy. Rather than allocating such an unnecessarily wide bandwidth, this article presents a controller that regulates the spectral resources adaptively to meet the desired reference accuracy while minimizing the total occupied bandwidth. The controller continuously monitors the statistical parameters of the received signal, such as signal-to-noise ratio, in a perception stage and adapts the spectral characteristics of the transmitted waveform in an action stage. The adaptive action, based on a Cramer–Rao lower bound analysis, maintains the signal statistical characteristics below a specified bound to maintain high coherent gain. Experimental results demonstrate the ability to maintain ranging standard deviation of 1.5 mm (standard deviation of time delay estimates equals $10^{-11}$ s), which yields 90% or more of the possible achievable coherent gain at carrier frequencies up to 13.33 GHz.

Sea-surface small target detection is always a difficult problem in high-resolution maritime ubiquitous radars for complex characteristics of sea clutter, weak target returns, and diversity of targets. Multiple features extracted from radar returns in different domains have ability but not enough to solely distinguish radar returns with target from sea clutter. Joint exploitation of multiple features becomes the key to improve detection performance. In this article, the K-nearest neighbor (KNN) algorithm and anomaly detection idea are cooperated to develop a novel sea-surface target detection method in the feature space spanned by the eight existing salient features. The detection is realized by the anomaly detection followed by a specially designed KNN-based classifier with a controllable false alarm rate. In the anomaly detection, a decision region is determined by the hyper-spherical coverage of the training set of sea clutter that is sufficient and ergodic in the feature space. The KNN-based classifier is designed based on the training sample set of sea clutter and the training sample set of simulated target returns plus sea clutter that is sufficient but nonergodic, by joint usage of feature weighting, neighbor weighting, and distance weighting. The novel method is validated by the two open and recognized IPIX and CSIR radar databases for sea-surface small target detection. The results show that it provides significant performance improvement in comparison with the existing multiple-feature-based detection methods, owing to the fact that the novel method avoids the dimension restriction and feature compression loss in the existing methods.

Recognition of multifunction radar (MFR) is an open problem in the field of electronic intelligence. Parameters of MFR pulses are generally agile and difficult to distinguish statistically. A prospective way to realize credible MFR recognition is mining and exploiting more distinguishable high-dimensional patterns buried in pulse groups, which may be designed for implementing infrequently used radar modes such as target tracking. A high-dimensional pattern is defined according to the agile range and switching law of sequential pulse repetitive intervals within a pulse group. This article establishes deep recurrent neural networks (RNN) to discriminate and coarsely cluster different pulse groups hierarchically with respect to their sequential structures. Afterwards, RNN-based classifiers are trained to extract and exploit features within different pulse group clusters. Distinct degrees of confidence are then attached to these classifiers to indicate the discriminabilities of the corresponding pulse group clusters. The pulse group clustering and classifying models are finally cascaded to form an integrated classification model, which mines distinguishable patterns from sequentially arriving pulse groups of the same radar and accumulate them to realize MFR recognition. Simulation results demonstrate the much improved performance of the proposed method over existing counterparts in different scenarios.

Recent developments in electronics have pushed miniaturized satellites to the femto-scale, with masses between 10 and 100 g. Although femtosatellites have been proven as a feasible concept, most designs are limited in mission capacity and lifetime due to the lack of environmental protection and onboard propellant. In this article, a novel concept for femtosatellites for earth remote sensing is proposed. In particular, a swarm of femtosatellites are used as elements of a sparse array in orbit to receive radar echoes. They also feature active orbit control enabled by solar radiation pressure to extend their lifetime. A simple active orbit control algorithm has been demonstrated. A mission concept based on a sun-synchronous circular orbit is proposed to maximize the benefit for both earth remote sensing and active orbit control. A synthetic aperture radar mission has been used to characterize their performance.

Shipborne or airborne maritime surveillance radars at scan mode work at a complex scene consisting of land, sea, and islands. Sea–land segmentation provides a two-class classification of sea clutter versus ground clutter to the surveillance scene as a precondition of adaptive target detection. It is a difficult problem because only a few coherent pulses are available at scan mode. Moreover, due to moving radar platforms and wide dynamic range of ground and sea clutter in power, average amplitude, Doppler offset, and initial phase of radar, the returns vector fails to distinguish sea clutter and ground clutter. In this article, a similarity measure of two radar returns vectors, which is invariant to amplitude, Doppler offset, and initial phase, is constructed, which is closely relevant to the Doppler bandwidth of a returns vector. Based on the similarity measure, a K-nearest neighbor classifier is proposed to yield a pixel-level sea–land segmentation of the scene. Further, the morphological filtering is operated on the pixel-level segmentation to obtain a region-level segmentation. Moreover, the discrete Frechet distances of the main boundaries in successive scan periods are used to assess segmentation quality. The proposed method is verified by measured data from an airborne radar and an island-based radar. The results show that it behaves better than the methods using thresholding phase linearity degree of radar returns and the support vector machine, and back propagation neural network in a three-dimensional feature space.

We present an ultrashort range IEEE 802.11ad-based automotive joint radar-communications (JRC) framework, wherein we improve the radar's Doppler resilience by incorporating Prouhet–Thue–Morse sequences in the preamble. The proposed processing reveals detailed microfeatures of common automotive objects verified through extended scattering center models of animated pedestrian, bicycle, and car targets. Numerical experiments demonstrate 2.5% reduction in the probability of false alarm at low signal-to-noise-ratios and improvement in the peak-to-sidelobe level dynamic range up to Doppler velocities of $\pm 144$ km/h over conventional 802.11ad JRC.

In modern antenna systems, beam broadening of subarrayed arrays provides continuous coverage of a wide angular extent in a cost-effective manner. While many methods have been published that address beam broadening of traditional (non-subarrayed) arrays, there is a knowledge gap in the published literature with respect to efficient beam broadening of contiguous uniform subarrayed arrays. This article presents efficient methods for beam broadening of contiguous uniform subarrayed arrays where the excitation of each element is not individually controlled, but the elements of the array are grouped together as subarrays to have the same element excitations. Particularly, this article focuses on phase-only optimization to preserve maximum power output. Three modified iterative Fourier transform (IFT) methods and one genetic algorithm (GA) are presented to efficiently search the vast solution space of possible phase settings for a solution that satisfies the desired broadened pattern. These methods are evaluated on idealized $1 \times 40$ and $1 \times 80$ linear arrays with five element subarrays and $40 \times 40$ and $80 \times 80$ element rectangular arrays with $5 \times 5$ element subarrays. The proposed modified IFT methods are found to be faster than the GA approach, while the GA approach only offers a few percentage points of better effectiveness.

Lots of detectors for high resolution range profile (HRRP) data are mainly based on the intensities of target echoes. When the intensities of outliers are as large as those of targets of interest, some false alarms will emerge during the detection stage. In this article, we propose a rejection algorithm for HRRP data between the detection stage and recognition stage to eliminate the false alarms occurring during the detection stage. In this method, the intensities and positions of the dominant scatterers are extracted as a joint feature. Then, the feature is used to develop the K-center one-class classifier based on Hausdorff distance. Experimental results on the measured HRRP data indicate that the proposed rejection algorithm can eliminate the false alarms remarkably well, and keep most of the target data as well. Meanwhile, the proposed method has better performance under different values of signal-to-noise ratio and different values of parameters than some other rejection methods.

Discriminant analysis is a technique used in statistics and machine learning to separate two or more classes of objects or events. We introduce linear, quadratic, and mixture discriminant analysis methods into radar signal classification. However, the selection of an appropriate discriminant analysis method can be difficult and no comparison study of these discriminant analyses for radar signal classification can be found in the open literature. This article presents a theoretical analysis and practical comparison of these three discriminant analysis methods for radar signal classification. The theoretical analysis is derived from Bayes’ theorem. The practical comparison study is performed on a dataset consisting of five radars emissions. The advantages and drawbacks of each discriminant analysis method are highlighted. This study demonstrates that quadratic discriminant analysis (QDA) is predominantly a better method for radar signal classification using our radar dataset. On average, it demonstrates 95%, 93%, and 86.6% classification accuracy for three, four, and five radar emitters in our radar dataset, respectively. Linear discriminant analysis (LDA) achieves on average 88.7%, 84.9%, and 79.2% classification accuracy for three, four, and five radar emitters, respectively. Mixture discriminant analysis (MDA) also achieves the same classification performance as QDA. Theoretical analysis shows that both LDA and QDA are a special case of MDA and that MDA can set up more decision boundaries than LDA and QDA if the feature distribution in the dataset is an ensemble of Gaussian distributions. Therefore, MDA is recommended for radar signal classification.

Deep neural networks have recently received a great deal of attention in applications requiring classification of radar returns, including radar-based human activity recognition for security, smart homes, assisted living, and biomedicine. However, acquiring a sufficiently large training dataset remains a daunting task due to the high human costs and resources required for radar data collection. In this article, an extended approach to adversarial learning is proposed for generation of synthetic radar micro-Doppler signatures that are well adapted to different environments. The synthetic data are evaluated using visual interpretation, analysis of kinematic consistency, data diversity, dimensions of the latent space, and saliency maps. A principle-component analysis-based kinematic-sifting algorithm is introduced to ensure that synthetic signatures are consistent with physically possible human motions. The synthetic dataset is used to train a 19-layer deep convolutional neural network to classify micro-Doppler signatures acquired from an environment different from that of the dataset supplied to the adversarial network. An overall accuracy of 93% is achieved on a dataset that contains multiple aspect angles (0$^{\circ }$, 30$^{\circ }$, and 45$^{\circ }$ as well as 60$^{\circ }$), with 9% improvement as a result of kinematic sifting.

The recognition of the interference environment for adaptive radar detection is addressed in this article. Typically, detectors are designed in one specific scenario which may not be appropriate for the varying interference environment, especially for the airborne and space-based radar system. In this article, the considered recognition task is cast in terms of multiple hypothesis tests and the theory of model order selection (MOS) techniques are exploited to devise suitable decision rules. The interference environments are divided into homogeneity, partial homogeneity, and spherically invariant random process. Three MOS techniques, namely, the Akaike information criterion (AIC), generalized information criterion, and corrected AIC, are adopted. At the analysis stage, illustrating examples for the influence of the environment model parameters on the recognition accuracy of the MOS rules are presented. Numerical experiments show the AIC rule has the most robust recognition performance.

Motivated by the problem of radar target recognition, we develop a label-aided factor analysis (LA-FA) model for statistical modeling of high-resolution range profile (HRRP) under the prerequisite that the HRRP data are Gaussian distributed. The LA-FA model is the extension of the multitask learning-based factor analysis (MTL-FA) model, which is mainly applied to the recognition problem with small training data size. Compared to the MTL-FA model, our LA-FA model introduces the discrete class labels via Sigmoid-Bernoulli hierarchy to restrict the learning of model parameters, which offers the potential to enhance the separability of statistical models from different classes, thus beneficial to the improvement of recognition capability. In addition, since the noise level of a test sample is usually different from those of the training samples in the real application, we introduce a noise-robust modification method for Gaussian-based models. The proposed modification method is implemented by updating the noise level parameter of the statistical model according to the estimated signal-to-noise ratio (SNR) of test HRRP. Experiments on measured HRRP data demonstrate the better recognition performance of the LA-FA model with limited training data and our noise-robust model modification method under low test SNR. Especially, when there are 20 training HRRP samples per frame, the recognition rate of our LA-FA model is 7% higher than that of the MTL-FA model, and moreover, the recognition accuracy of the noise-robust LA-FA model is 3% higher than that of the LA-FA model without modification under the condition of SNR = 15 dB.

Aiming at the signal reconstruction problem for the conventional narrowband radar system, we propose a new statistical compressive sensing (SCS) method to achieve the reconstruction of superresolution time-frequency spectrum from the corrupted time-domain measurement. The proposed method assumes that the signal obeys complex Gaussian distribution and develops a hierarchical Bayesian model. Variational Bayesian expectation maximization (VBEM) is used to perform inference for the posterior distributions of the model parameters. In order to fully exploit the superresolution characteristics of reconstructed spectrum, a novel superresolution time-frequency feature vector is extracted for subsequent classification of ground moving targets, i.e., walking person and a moving wheeled vehicle. Experimental results on measured data show that the proposed reconstruction method can obtain good reconstruction results and the superresolution feature has good classification performance for human and vehicle targets.

The backscattered signals of objects under spinning motion or with rotating parts provide very rich information that can be used for classification tasks, parameter extraction, etc. Obtaining such information from noncooperative objects with an unknown target-aspect is often a complicated task with a monostatic configuration. A multistatic radar on the other hand, can exploit the spatial diversity to extract the information from the time–frequency representations obtained from multiple aspect-angles. In this paper, we propose a tomographic approach for characterizing spinning objects in terms of their shape, size, and rotation parameters using a narrow-band multistatic radar. A two-dimensional image is reconstructed after a full rotation period using tomographic methods that allows not only to estimate the shape of the target but also the rotation parameters and the dimensions of the object. This is done very efficiently by combining the tomographic images from different aspect-angles on the transformed log-polar space, instead of the time–frequency representations. Simulations and measurements were conducted for the proof of concept. The measurement results with a simple target and a continuous wave K-band radar show errors below 3 $^{\circ }$ for the orientation estimation and below 5% for the estimation of the object's diameter.

In previous works, methods were explored for position estimation utilizing satellite-borne signals of opportunity, mainly the global positioning system (GPS). The GPS signal was exploited for use in a multistatic passive coherent location (PCL) system. The GPS signal is especially attractive for PCL applications because of the native capability to produce position and velocity estimation. This paper examines the signal specifications for PCL implementation and explores the potential limitations of the proposed solutions. GPS specific methods are developed for multistatic PCL velocity estimation in a three-dimensional plane. The method developed is combined with previously completed work of GPS PCL position estimation for a complete system design in range and Doppler. The PCL system is evaluated against conventional GPS position estimation and velocity estimation and proves to have comparable metrics of performance. Analysis and simulation are performed for verification and validation of the developed methods.

Automatic dependent surveillance-broadcast (ADS-B) is an emerging air traffic surveillance technology that overcomes the limitations of today's radar technology and enhances the air traffic control. To provide authenticity, integrity in ADS-B and to improve the computation, communications efficiency, in this paper, we propose a new, efficient, and secure pairing-free ADS-B authentication scheme with Batch Verification in ID-based framework. The proposed scheme is proven secure and is more efficient than the existing schemes.

In this study, a new passive method suitable for geolocating low probability of intercept (LPI) radars is introduced. The method uses two electronic support receivers placed on a fast moving platform (e.g., an airplane or a satellite). The proposed method has a high processing gain, which makes it highly suitable for very weak LPI signals. The method processing gain was analytically derived and illustrated with simulation to determine if the proposed method can detect LPI radars in much lower SNRs compared to regular time–frequency LPI detection methods. Also, the resolution of the proposed method in both range and cross-range directions in radar location finding was analyzed. The results showed that the method is capable of radar location finding for complex radars and in complicated electromagnetic environments.

The effect of the radar skin return from the platform on which a cross-eye jammer is mounted is significant in many practical cross-eye jamming scenarios. However, all published analyses of skin-return affected cross-eye jamming have significant limitations. These limitations are addressed by deriving equations for the distribution of the cross-eye gain in the presence of skin return. The values of these results are demonstrated by using them to gain insight into how skin return affects cross-eye jamming.

Using low-frequency (UHF to L-band) ultrawideband synthetic aperture radar (SAR) technology for detecting buried and obscured targets, e.g., bomb or mine, has been successfully demonstrated recently. Despite promising recent progress, a significant open challenge is to distinguish obscured targets from other (natural and manmade) clutter sources in the scene. The problem becomes exacerbated in the presence of noisy responses from rough ground surfaces. In this paper, we present three novel sparsity-driven techniques, which not only exploit the subtle features of raw captured data, but also take advantage of the polarization diversity and the aspect angle dependence information from multichannel SAR data. First, the traditional sparse representation-based classification is generalized to exploit shared information of classes and various sparsity structures of tensor coefficients for multichannel data. Corresponding tensor dictionary learning models are consequently proposed to enhance classification accuracy. Finally, a new tensor sparsity model is proposed to model responses from multiple consecutive looks of objects, which is a unique characteristic of the data set we consider. Extensive experimental results on a high-fidelity electromagnetic simulated data set and radar data collected from the U.S. Army Research Laboratory side-looking SAR demonstrate the advantages of proposed tensor sparsity models.

Frequency-modulated continuous waveform signals are of great interest in automotive radar, perimeter radar, and wide-area surveillance radar applications. For automotive radars and wide-area surveillance radars, which generally utilize high Doppler frequency resolution for target recognition, the important potential target such as a walking person cannot be viewed as a point-like target due to different velocities of the reflection points of a pedestrian. The echoes of such targets have extended Doppler spectrum, and they appear as horizontal lines in the range-Doppler plane, rather than a single point considered in the traditional cases. However, the existing approaches do not make full use of the Doppler-spread characteristic to enhance the detection performance of pedestrians. In this paper, we will show how this characteristic can be used to enhance the detection performance for Doppler-spread targets. The main goal of this paper is pedestrian detection in wide-surveillance radar applications and the proposed approach can also be applied in automotive radar applications. Motivated by the Hough transform, the energies of most Doppler bins corresponding to the target will be accumulated to enhance the detection performance. The questions of how many and which cells should be accumulated in practical applications are discussed in detail. The proposed approach is based on the ordered statistical-constant false alarm rate but tailored to Doppler-spread targets. The detection performance of this algorithm is derived analytically, and verified via practical wide-area surveillance radar signals.

Joint radar-communications dual function has drawn lots of attention since it can make a better use of the scarce wireless frequency resources and expensive hardware platforms. In case of joint radar-communications signal co-use, many communication sequences have poor range sidelobes and thus are not very suitable for the radar function. In this paper, we present a single carrier joint radar-communications method operating in the pulsed radar mode. Digital communication sequences are first partitioned into blocks which are then mapped to digital phase-coded sequences, like binary phase-shift keying (BPSK) and quadrature phase-shift keying (QPSK) sequences. The phases of the digital sequences are perturbed a bit such that certain degrees of freedom are available to optimize for lower range sidelobes. Insignificant phase perturbation will be deemed as phase noise by a communication receiver and then phase codes can be correctly decoded; in radar-processing channels, range compression are performed with known and optimized phase perturbation such that low-range sidelobes are obtained. An implementation scheme is presented. Numerical results with BPSK and QPSK sequences indicate that little phase perturbation can significantly drop the range sidelobe level but will insignificantly rise the bit error rate.

We propose a millimeter-wave joint radar-communications (JRC) system comprising a bi-static automotive radar and vehicle-to-vehicle communications. We study the applicability of known phase-modulated-continuous-wave (PMCW) and orthogonal-frequency-division-multiple-access waveforms for bi-static-JRC. In both cases, we design multiplexing strategies to ensure the parameter identifiability, derive JRC statistical bounds and numerically demonstrate superior performance of proposed low-complexity JRC super-resolution algorithms over conventional two-dimensional fast Fourier transform/MUltiple SIgnal Classification.

Both economic incentives and policy trends motivate the study of spectrum sharing between radar and wireless communications as a means to mitigate spectrum scarcity. This paper proposes a novel, temporal signal processing algorithm for a radar to mitigate interference from cyclostationary, linearly modulated digital communications (LMDC) interference such as orthogonal frequency division multiplexing (OFDM) or code division multiple access (CDMA). The proposed algorithm has the form of a novel whitening filter followed by a matched filter, and as such it optimizes the statistical deflection among the class of all linear detection filters. The derived cyclostationary whitening filter has equivalent mathematical representations as: 1) the form of a multiuser detector followed by an interference canceler, and 2) a frequency-shift (FRESH) filter. Performance results indicate that the derived cyclostationary whitener can produce signal-to-interference-plus-noise ratio (SINR) improvements ranging from negligible to up to $\text{20 dB}$ for the parameters used in this study. The cyclostationary whitener leads to the greatest SINR improvements when the interference-plus-noise is dominated by only a few LMDC signals, such as might occur in a near-far scenario.

The radar echoes of space debris sometimes have drawbacks of inadequate pulse samples, self-occlusion, and data missing, while the state-of-art radar imaging algorithms become unproductive. To tackle this problem, this paper proposes a novel method for three-dimensional (3-D) imaging of space debris. This iterative algorithm terminates automatically and can recover the intensity of space debris radar cross section, making it practical in real-world applications. Unlike the 3-D imaging methods based on multistatic inverse synthetic aperture radar, the proposed method is applicable to monostatic ISAR.

We address the problem of designing a transmit beampattern for multiple-input multiple-output (MIMO) radar considering coexistence with wireless communication systems. The designed beampattern is able to manage the transmit energy in spatial directions as well as in spectral frequency bands of interest by minimizing the deviation of the designed beampattern versus a desired one under a spectral constraint as well as the constant modulus constraint. While unconstrained beampattern design is straightforward, a key open challenge is jointly enforcing the spectral constraint in addition to the constant modulus constraint on the radar waveform. A new approach is proposed in this paper, which involves solving a sequence of constrained quadratic programs such that constant modulus is achieved at convergence. Furthermore, we show that each problem in the sequence has a closed form solution leading to analytical tractability. We evaluate the proposed beampattern with interference control (BIC) algorithm against the state-of-the-art MIMO beampattern design techniques and show that BIC achieves closeness to an idealized beampattern along with desired spectral shaping.

An innovative moving target detection approach termed as Beam-Doppler Image Feature Recognition (BDIFR) is introduced based on the distinction between moving target and ground clutter features in the beam-Doppler domain. A novel minimum-distance-based region-growing method is developed for radar target feature extraction and detection. The proposed BDIFR algorithm is advantageous over conventional space-time adaptive processing in detecting ground moving targets in inhomogeneous clutter environments since it does not require secondary training data for clutter estimation.

The problem of centralized adaptive constant false alarm rate (CFAR) detection with registration errors in multistatic radar is considered. When the observation data from different radar sites are not properly aligned after transformed into a common coordinate system, registration errors occur. This can severely degrade the target detection performance and positioning accuracy of multistatic radar. In this paper, we focus on the design of adaptive detectors that are applicable for centralized target detection with registration errors in multistatic radar. A maximum likelihood estimation-based generalized likelihood ratio test detector and a maximum a posteriori estimation-based generalized likelihood ratio test detector are developed. Both detectors possess the CFAR property with regard to the unknown statistics of the background noise. Accordingly, the performance of the two proposed detectors are analyzed.

Ground moving target indicator (GMTI) radar processing attempts to distinguish between radar returns emanating from moving targets and stationary ground clutter. The task is confounded by the relative motion between the radar platform and the scene, as well as by the strength of clutter returns. Techniques such as space-time adaptive processing require an unknown interference covariance describing clutter, jammers, and thermal noise. The covariance is estimated from training data not under test, but, heterogeneous, contaminated, or limited training data degrade the covariance estimate and reduce the detection performance. State-of-the-art techniques for interference covariance estimation reduce the required amount of training data by imposing assumed structure on the covariance matrix. Here, a Bayesian signal model is adopted for jointly estimating targets and clutter in a single cell under test, allowing GMTI processing without training data. The approach incorporates the knowledge of an approximate digital elevation map, platform kinematics (velocity, crab angle, and antenna spacings), and the belief that moving targets are sparse in the scene. Low-complexity computation with the Bayesian model is enabled by recent algorithm developments for fast inference on linear mixing models. Results from the KASSPER I dataset show improved detection performance compared to existing techniques using scores or even hundreds of training bins.

We present a Xampling-based technology enabling interference-free operation of radar and communication systems over a common spectrum. Our system uses a recently developed cognitive radio (CRo) to sense the spectrum at low sampling and processing rates. The Xampling-based cognitive radar (CRr) then transmits and receives in the available disjoint narrow bands. Our main contribution is the unification and adaptation of two previous ideas—CRo and CRr—to address spectrum sharing. Hardware implementation shows robust performance at SNRs up to –5 dB.

In this paper, a sparsity-driven method of micro-Doppler analysis is proposed for dynamic hand gesture recognition with radar sensors. First, sparse representations of the echoes reflected from dynamic hand gestures are achieved through the Gaussian-windowed Fourier dictionary. Second, the micro-Doppler features of dynamic hand gestures are extracted using the orthogonal matching pursuit algorithm. Finally, the nearest neighbor classifier is combined with the modified Hausdorff distance to recognize dynamic hand gestures based on the sparse micro-Doppler features. Experiments with real radar data show that the recognition accuracy produced by the proposed method exceeds 96% under moderate noise, and the proposed method outperforms the approaches based on principal component analysis and deep convolutional neural network with small training dataset.

Recent changes in U.S. policy call for spectrum sharing between radar and wireless communications. In order to inform choices of protection criteria for a radar sharing spectrum with communications systems, this paper models communications interference and its effects on a cell-averaging adaptive-threshold radar detector. First, we propose a statistical model for communications interference based on existing models and original simulations, and we find support for both Gaussian and non-Gaussian models, depending on the application. Then, we assess the impact of communications interference on radar detection. Our results suggest that when interference is not well modeled as white Gaussian noise, mean interference-to-noise ratio (INR) is not sufficient to characterize interference effects on radar, and additional interference characteristics such as kurtosis/impulsiveness and coherence time can significantly impact radar performance. We also find degradation in radar performance in relatively low-INR Gaussian interference (e.g., about $-6$ to $-\text{2 dB}$ mean INR at the output of the matched filter).

Multiple-input multiple-output (MIMO) radar is an enabling technique for high-resolution imaging, which is especially useful for near-field electromagnetic scattering diagnosis of complex targets. Among others, high sidelobes and radar cross section (RCS) calibration uncertainty are the major challenges for such applications, due to array nonuniformity, imperfect channels, and antenna pattern tapering. These shortcomings prevent a MIMO radar from obtaining high-quality images with enough dynamic range and RCS accuracy. In this paper, we develop a complete solution for these problems. A novel adaptive weighting technique is proposed, where the complex weights are optimized for exact amplitude and phase error calibration of a MIMO array and for azimuth sidelobe reduction. A MIMO filtered backprojection algorithm is developed for image formation with improved RCS calibration accuracy, where propagation path-loss, antenna pattern tapering, and phase distortion due to the near-field spherical wave front are compensated. Both indoor and outdoor field test results are presented to show the high-quality images obtained using the proposed techniques, demonstrating the applicability of a MIMO radar for diagnostic RCS imaging of complex targets.

Studies have demonstrated the usefulness of micro-Doppler signatures for classifying dynamic radar targets such as humans, helicopters, and wind turbines. However, these classification works are based on the assumption that the propagation channel consists of only a single moving target. When multiple targets move simultaneously in the channel, the micro-Dopplers, in their radar backscatter, superimpose thereby distorting the signatures. In this paper, we propose a method to detect multiple targets that move simultaneously in the propagation channel. We first model the micro-Doppler radar signatures of different movers using dictionary learning techniques. Then, we use a sparse coding algorithm to separate the aggregate radar backscatter signal from multiple targets into their individual components. We demonstrate that the disaggregated signals are useful for accurately detecting multiple targets.

This paper addresses the issue of beampattern synthesis with flexible magnitude response and low sidelobe in a frequency diverse array radar. It is formulated as an optimization problem in terms of $\ell _1$ -norm minimization, which encourages the sparsity of array pattern. An $\ell _1$ iterative phase compensation (IPC) technique is employed to transform the nonconvex constraint to phase compensation form. Then, to further reduce the sidelobe level, a reweighted $\ell _1$ -IPC algorithm is devised, in which a sparsity-enhanced scheme is utilized to convert the $\ell _1$ -norm to a new cost function that more closely resembles $\ell _0$ -norm. Numerical results are presented to demonstrate the effectiveness of the proposed approach in different scenarios.