The major clutter in a typical ground penetrating radar (GPR) B-scan image is the ground surface clutter. It can severely obscure or distort the subsurface target response, especially when the surface profile variation is large and the target reflection signal is overlapped with the surface clutter. In this paper, we propose a deep auto-encoder based method for reducing the rough surface clutter in GPR images. In our method, the rough surface clutter reduction is formulated as an anomaly detection problem. The rough surface region in a B-scan image is splitted into small patches which are used as the training data set to train a deep auto-encoder. The trained autoencoder captures the patterns of the rough surface patches. After training, the whole B-scan image is divided into small patches of the same size as the training patches and each of which is fed into the auto-encoder to compute an anomaly score. To reduce the clutter, we design a novel method that adopts a weighted sum approach to aggregate all patches based on their anomaly scores. For performance evaluation, simulation and laboratory tests are conducted, and the clutter-to-rest ratios (CRR) are computed to quantify the clutter reduction effectiveness. The CRR of the simulated B-scan image is reduced from 4.24 dB to -14.06 dB, and the CRRs of two measured B-scans are reduced from 3.71 dB to -20.45 dB, and from 8.62 dB to -8.07 dB, respectively.
In this paper, we present an experimental evaluation of recently proposed Supervised Reciprocal Filter approaches for the compression of OFDM-radar signals. The range-Doppler map is usually evaluated using a suboptimal batches algorithm, after fragmenting the signal in batches with length equal to the OFDM symbol. Using “OFDM fragmentation” requires symbol synchronization and sets constraints on the non-ambiguous Range-Doppler area of targets that can be detected with limited Signal-to-Noise Ratio (SNR) loss. Supervised Reciprocal Filters have been recently proposed to operate with batches of longer lengths than the OFDM symbol without requiring any synchronization. In this paper we extend the study to include the case of batches equal to a fraction of the OFDM symbol, which provides higher flexibility to adapt the processor to the range-Doppler scenario of interest. These filters have been shown to contain the large SNR losses obtained with a direct application of the Reciprocal Filter (RF) with the non-OFDM fragmentation. Moreover, they have been shown theoretically to preserve the benefits of the RF over the Matched Filter (MF) against the clutter-limited scenarios. To assess the performance of the Supervised Filter against a real scenario, an acquisition campaign has been carried out using the Sapienza experimental passive radar along the coast north of Rome, against a maritime traffic scenario, including non-cooperative vessels, as well as a cooperating small boat equipped with differential GPS positioning registration tools. The effectiveness of the proposed approaches is validated by applying them to experimental data from a PBR application exploiting DVB-T transmissions.
Ship detection in synthetic aperture radar (SAR) images is a hot pot in the remote sensing (RS) field. However, most existing deep learning (DL)-based methods only focus on the single-polarization SAR ship detection without leveraging the rich dual-polarization SAR features, which poses a huge obstacle to the further model performance improvement. One problem for solution is how to fully excavate polarization characteristics using a convolution neural network (CNN). To address the above problem, we propose a novel group-wise feature fusion R-CNN (GWFF R-CNN) for dual-polarization SAR ship detection. Different from raw Faster R-CNN, GWFF R-CNN embeds a group-wise feature fusion module (GWFF module) into the subnetwork of Faster R-CNN, which enables group-wise feature fusion between polarization features and multi-scale ship features. Finally, the experiments on the dual-polarization SAR ship detection dataset (DSSDD) demonstrate that GWFF R-CNN can yield a ~4.1 F1 improvement and a ~2.9 average precision (AP) improvement, compared with Faster R-CNN.
This paper proposes a novel passive nonlinear tag design, fabrication, and tests using a second-order harmonic system with various antenna configurations. The tags' antennas were fabricated with laser-cut pattern masks and conductive materials, and their performance was greatly enhanced by the multi-element design and Schottky diode's nonlinearity. The detectable distance of up to 6.7 ft of the tags was tested using a benchtop setup of a 7.9 GHz transmitter and a 15.8 GHz receiver antenna without any signal amplification. The lightweight and miniature tags have great potential for extensive applications in locating objects.
When radar detects a high-speed maneuvering target, not only will the phenomena of range migration (RM) and Doppler migration (DM) appear, but also the phenomenon of range ambiguity, which poses challenges to the traditional accumulation processing method. In this paper, we first establish the target echo model with range ambiguity based on the spatial geometric model. On this basis, we propose a coherent integration method based on the modulo generalized Radon Fourier transform (MGRFT). By performing the modulo addressing operation during the joint search of motion parameters, the proposed method can correct RM and DM and deal with the problem of trajectory breakage under range ambiguity so as to achieve the coherent integration of echo energy and effectively improve the signal-to-noise ratio (SNR). Finally, experimental results demonstrate the effectiveness of the proposed algorithm.
This paper considers target detection in distributed multi-input multi-output (MIMO) radar with non-orthogonal waveforms in non-homogenous clutter. We first present a general signal model for distributed MIMO radar in cluttered environments. To cope with the non-homogenous clutter and possible clutter bandwidth mismatch, the covariance matrix of the disturbance (clutter and noise) signal is modeled as a random matrix following an inverse complex Wishart distribution. Then, we propose three Bayesian detectors, including a non-coherent detector, a coherent detector, and a hybrid detector. The latter is a compromise of the former two, as it forsakes phase estimation needed by the coherent detector, but requires the samples within a coherent processing interval (CPI) to maintain phase coherence that is unnecessary for the non-coherent detector. Simulation results are presented to illustrate the performance of these Bayesian detectors and their non-Bayesian counterparts in non-homogeneous clutter when the clutter bandwidth is known exactly and, respectively, with uncertainty.
Dynamic spectrum sharing between airborne radars and 5G cellular networks has the potential for granting additional RF spectrum to cellular networks while preserving the performance of airborne radars. In the case of an airborne radar with a predictably rotating antenna, a spectrum sharing controller can use estimates of the radar's location and beam orientation to anticipate and mitigate RF interference events over a large geographic area. However, localization of the radar is complicated by airborne radar's relatively narrow beamwidth and timevarying waveform. We introduce the Rotating Beam Time-of-Arrival (RB-TOA) algorithm to jointly estimate the radar's location and antenna main beam orientation. Each RF sensor is coarsely time-synchronized and measures the peak of the received signal envelope over each rotation interval to estimate when the radar's main beam maximally couples with the sensor's antenna; these time estimates are then combined at a sensor fusion server and the radar's main beam orientation and location are jointly solved using a gradient descent algorithm. We show that the RB-TOA algorithm rapidly converges to a geolocation accuracy that is 50x better than the performance of a two-antenna angle-of-arrival algorithm (AoA) for the same number of sensors.
The emerging 4D-imaging automotive MIMO radar sensors necessitate the selection of appropriate transmit wave-forms, which should be separable on the receive side in addition to having low auto-correlation sidelobes. TDM, FDM, DDM, and inter-chirp CDM approaches have traditionally been proposed for FMCW radar sensors to ensure the orthogonality of the transmit signals. However, as the number of transmit antennas increases, each of the aforementioned approaches suffers from some drawbacks, which are described in this paper. PMCW radars, on the other hand, can be considered to be more costly to implement, have been proposed to provide better performance and allow for the use of waveform optimization techniques. In this context, we use a block gradient descent approach to design a waveform set for MIMO-PMCW that is optimized based on weighted integrated sidelobe level in this paper, and we show that the proposed waveform outperforms conventional MIMO-FMCW approaches by performing comparative simulations.
Advanceddriver assistance systems (ADASs) and autonomous vehicles rely on differenttypes of sensors, such as cameras, radar, ultrasonic, and LiDAR to sense thesurrounding environment. Compared with the other types of sensors,millimeter-wave automotive radar has advantages in terms of low hardware costand reliable object detection under poor weather conditions, such as snow,rain, or fog, and doesn't suffer from light condition variations, such asdarkness. High-resolution radar bird's-eye-view (BEV) obtained from radarrange-azimuth spectra through a polar-to-Cartesian coordinate transformcontains targets’ geometric information that can be learned by deep neuralnetworks for object detection. Compared to radar point clouds, there is noinformation loss in radar BEV. Unlike RGB images, radar BEVs are single-channelgrayscale images with unique characteristics such as inconsistent resolutionand SNR. Therefore, directly implementing an image-based object detectionnetwork is not an optimal solution for object detection using radar BEV. Wepropose a Temporal-fusion, Distance tolerant single stage object detectionNetwork, termed as, TDRadarNet, to robustly detect vehicles up to 100 metersunder various driving scenarios. DRadarNet leverages historical radar frames toexploit temporal features and separates far and near fields to addressinconsistent resolution in radar frames. With qualitative and quantitativeresults, we show that TDRadarNet achieves 68.9% in precision and 66.8% inrecall, and 67.8% in F1-score, which outperforms the state-of-the-artimage-based object detection networks by 10.6%, 17.1%, and 14.1%.
2D-ISAR produces images that strictly depend on the geometry of the whole radar-target system and on the relative motion between radar and target. This poses some limits on the use of Automatic Target Recognition (ATR) systems. To overcome this issue, 3D point clouds as a result of 3D-ISAR imaging were proposed as a more complete and reliable representation of the target. Since the acquisition system will output an unknown number of points in a random order, the chosen classifier must be able to process a variable number of input elements to correctly classify the target. After a brief presentation of the state of the art about the 3D classification problem, the architecture of Point Cloud Transformer (PCT) is introduced. PCT is trained and tested on an ad-hoc generated 3D dataset, which in this preliminary experiment contains three different target types: cars, tanks and military trucks. The goal of this work is to show how the transformer is able to correctly manage the recognition of targets, even if the point clouds are made by few points. Lastly, the trained network is tested on some real data.
Spectrally shaped forms of random frequency modulation (RFM) radar waveforms have been experimentally demonstrated for a variety of implementation approaches and applications. Of these, the continuous-wave (CW) perspective is particularly interesting because it enables the prospect of very high signal dimensionality and arbitrary receive processing from a range/Doppler perspective, while also mitigating range ambiguities by avoiding repetition. Here we leverage a modification to the constant-envelope orthogonal frequency division multiplexing (CE-OFDM) framework, which was originally proposed for power-efficient communications, to realize a nonrepeating FMCW radar signal that can be represented with a compact parameterization, thereby circumventing memory constraints that could arise for some applications. Experimental loopback and open-air measurements are used to demonstrate this waveform type.
A new method to jointly detect and classify drones using a moving surveillance radar system ('radar on-the-move') and computer vision is presented. While most conventional counter-drone radar-based techniques focus on time-frequency distributions to obtain classification features, such approaches are limited in volumetric spatial coverage. To compensate for this, surveillance radars that offer full spatial coverage are used, but the determination of the best detection and classification approach to be applied on the resulting data is still an open challenge. In this paper a framework is proposed that combines deep learning approaches from computer vision, specifically the You Only Look Once (YOLO) network, with data from the moving surveillance radar produced by Robin Radar Systems B.V. This framework allows to jointly detect and label targets based on range-Doppler images generated in real-time. The method is validated on experimental data, with preliminary results on a small dataset showing precision, recall, mean average precision (mAP@0.5) and Area Under Curve (AUC) of over 99%.
Designing radar waveforms with notched spectral regions can mitigate mutual interference with other proximate RF users. However, this capability comes at the cost of degraded range-Doppler sidelobe performance. To evaluate the limitations of correlation-based processing, the null-constrained power spectral density that minimizes correlation sidelobe levels is determined for comparison with waveform and pulse compression filter design methods. Existence of the least-squares (LS) global optimum indicates a fundamental dynamic range limitation for notched power spectra (notwithstanding further receive compensation or range resolution spoiling). Recent work investigated spectrally notched random FM (RFM) waveform design where ad-hoc tapering was incorporated into the null shape as a heuristic means of reducing range sidelobes. Here, waveforms designed according to the optimal null-constrained spectral template are demonstrated to have improved sidelobe performance after pulse compression and slow-time processing. Further, because these waveforms are designed according to the LS optimal spectral template, application of the LS mismatched filter provides additional sidelobe reduction (toward the global limit) with minimal mismatch loss.
The wavelength used for illumination dictates the scale of the mechanisms that interact with the incident electromagnetic (EM) energy. We model the synthetic Aperture Radar Image of a target as a superposition of the returns from scattering mechanisms that depend on the wavelength of the illuminating waveform and the viewing angle. In this work, we present a method to jointly model the scattering responses of the target over a wide aperture of measurements and a wide swath of frequencies spanning the C to X Band. Specifically, we estimate the location of the scattering centers and their azimuth-dependent responses normalized by the wavelength, jointly for low and high bands. We verify the validity of the proposed model using simulated data from a backhoe and Civilian vehicle data domes dataset over two non-overlapping frequency bands centered at 7GHz and 12 GHz.
Smart noise jamming is an emerging barrage jamming and plays an essential role in electronic countermeasure (ECM). To improve the interference effectiveness of smart noise, in this paper, an exploratory deep deterministic policy gradient (EDDPG) algorithm is proposed to continuously adjust the multi-step jamming power. Firstly, multi-step jamming power adjustment is modeled as a Markov decision process (MDP). Subsequently, average jamming-to-signal ratio (JSR) at multifunctional radar (MFR) receiver is chosen as the evaluation indicator to assess the performance of multi-step noise. Moreover, the principle of noise jamming power allocation is analyzed, and a reinforcement learning framework is developed to continuously adjust multi-step jamming power. Finally, numerical results are provided to verify the validity of the proposed method.
In this paper, we investigate the performance of matched filtering (MF) for massive MIMO radar with one-bit ADCs. Firstly, we show that in the context of white Gaussian noise, the MF output of one-bit quantized received signals of mas-sive MIMO radar is asymptotically Gaussian. Then, statistical characteristics, including mean and covariance matrix, of the MF output are derived, respectively. More importantly, using the fact that massive MIMO radar has a large number of measurements (i.e., the number of samples in space/frequency/time domains), we provide approximate probabilistic distribution of the MF output, which is capable of making the signal processing algorithms of one-bit MIMO radar low in complexity. Moreover, based on the above approximations, the performance gap between one-bit and traditional infinite-bit MIMO radars is mathematically derived. Finally, from the perspective of target detection and beamforming, representative simulations are conducted to demonstrate the performance of massive MIMO radar with one-bit ADCs.
Dual-function radar-communication (DFRC) systems offer high spectral, hardware and power efficiency, as such are prime candidates for 6G wireless systems. DFRC systems use the same waveform for simultaneously probing the surroundings and communicating with other equipment. By exposing the communication information to potential targets, DFRC systems are vulnerable to eavesdropping. In this work, we propose to mitigate the problem by leveraging directional modulation (DM) enabled by a time-modulated array (TMA) that transmits OFDM waveforms. DM can scramble the signal in all directions except the directions of the legitimate user. However, the signal reflected by the targets is also scrambled, thus complicating the extraction of target parameters. We propose a novel, low-complexity target estimation method that estimates the target parameters based on the scrambled received symbols. We also propose a novel method to refine the obtained target estimates at the cost of increased complexity. With the proposed refinement algorithm, the proposed DFRC system can securely communicate with users while having high-precision sensing functionality.
This paper investigates the feasibility of one of noise radar's supposed main benefits, the proposed low probablity of intercept (LPI) due to the wideband pseudorandom noise signal being undetectable. The performance of both the noise radar and the intercepting receiver are first studied theoretically through stochastic signal analysis with multi-channel cross-correlating receivers, accompanied by simulation studies with realistic noise-modulated FM-signals and OFDM noise signals. The results are confirmed by a real measurement of an operating noise radar with an intercepting two-channel receiver both from mainlobe and sidelobe directions. It is shown that the interceptor with a correlation receiver has a significant SNR benefit in detecting the signal of the noise radar compared to the noise radar's own detection, even though the radar can utilize the exact knowledge of the transmit signal. This leads to 100-fold detection ranges compared to the radar's operation range rendering the supposed undetectability invalid. The main stealth benefit of noise radars is concluded to be the interceptor's inability to accurately predict the purpose of the detected wideband noise waveform.
Dual-function radar-communication (DFRC) design is a promising approach for solving the challenging spectrum congestion problem. This paper considers joint antenna selection and digital beamforming design for a DFRC system that serves multiple multicast communication groups and, meanwhile, performs sensing. The dual-function transmit design is cast as maximizing the minimum target illumination power in multiple target directions by jointly selecting the antennas and designing the beamformers subject to a lower bound on the signal-to-interference-plus-noise ratio (SINR) for the communication users and an upper bound on the clutter power at each clutter scatterer. The resulting optimization formulation is a mixed integer programming problem that is solved with a penalized sequential convex relaxation scheme along with semidefinite relaxation (SDR). Numerical results verify the effectiveness of the proposed DFRC scheme and the associated algorithm.
