Search Results (23)

To address the high computational cost and significant resource consumption of radar Doppler-based target recognition, which limits its application in real-time embedded systems, this paper proposes a lightweight CNN (Convolutional Neural Network) approach for radar target identification. The proposed approach builds a deep convolutional neural network using range-Doppler maps, and leverages data collected by frequency-modulated continuous wave (FMCW) radar from targets such as drones, vehicles, and pedestrians. This method enables efficient object detection and classification across a wide range of scenarios. To improve the performance of the proposed model, this study incorporates a coordinate attention mechanism within the convolutional neural network. This mechanism fine-tunes the network’s focus by dynamically adjusting the weights of different feature channels and spatial regions, allowing it to concentrate on the most informative areas. Experimental results show that the foundational architecture of the proposed deep learning model, RangDopplerNet Type-1, effectively captures micro-Doppler features from range-Doppler maps across diverse targets. This capability enables precise detection and classification, with the model achieving an impressive average recognition accuracy of 96.71%. The enhanced network architecture, RangeDopplerNet Type-2, reached an average accuracy of 98.08%, while retaining a compact footprint of only 403 KB. Compared with standard lightweight models such as MobileNetV2, the proposed architecture reduces model size by 97.04%. This demonstrates that, while improving accuracy, the proposed architecture also significantly reduces both computational and storage overhead.The deep learning model introduced in this study is specifically tailored for deployment on resource-constrained platforms, including mobile and embedded systems. It provides an efficient and practical approach for development of miniaturized low-power devices. Full article

The evolution of a tornadic supercell in Kansas on 24 May 2021 is documented from an analysis of data from the ground-based mobile RaXPol (Rapid-scan, X-band, Polarimetric) radar. A cyclonic tornado evolved from a single-vortex into a multi-vortex tornado. The formation and evolution of an anticyclonic tornado, which passed directly over the radar, is also documented, in addition to an anticyclonic, satellite vortex that moved along or just outside the outer edge of the cyclonic tornado. This study is noteworthy because there were both extensive radar and visual observations of the evolution of the tornadoes at close range. Full article

With the rapid expansion of Wind Parks (WPs), Wind Turbine Clutter (WTC) has become a significant challenge due to the interference it causes with data from next-generation Doppler weather radars. Traditional clutter detection methods struggle to strike a balance between detection accuracy and efficiency. This study proposes a deep learning model named WTC-MobResNet, which integrates the architectures of MobileNet and ResNet and is specifically designed for WTC detection tasks. The model combines the lightweight characteristics of MobileNet with the residual learning capabilities of ResNet, enabling efficient extraction of WTC features from weather radar echo data and achieving precise identification of WTC. The experimental results demonstrate that the proposed model achieves an ACC of 98.21%, a PRE of 97.52%, a POD of 98.99%, and an F1 score of 98.25%, outperforming several existing deep learning models in both detection accuracy and false alarm control. These results confirm the potential of WTC-MobResNet for real-world operational applications. Full article

Hand gesture recognition (HGR) is a convenient and natural form of human–computer interaction. It is suitable for various applications. Much research has already focused on wearable device-based HGR. By contrast, this paper gives an overview focused on device-free HGR. That means we evaluate HGR systems that do not require the user to wear something like a data glove or hold a device. HGR systems are explored regarding technology, hardware, and algorithms. The interconnectedness of timing and power requirements with hardware, pre-processing algorithm, classification, and technology and how they permit more or less granularity, accuracy, and number of gestures is clearly demonstrated. Sensor modalities evaluated are WIFI, vision, radar, mobile networks, and ultrasound. The pre-processing technologies stereo vision, multiple-input multiple-output (MIMO), spectrogram, phased array, range-doppler-map, range-angle-map, doppler-angle-map, and multilateration are explored. Classification approaches with and without ML are studied. Among those with ML, assessed algorithms range from simple tree structures to transformers. All applications are evaluated taking into account their level of integration. This encompasses determining whether the application presented is suitable for edge integration, their real-time capability, whether continuous learning is implemented, which robustness was achieved, whether ML is applied, and the accuracy level. Our survey aims to provide a thorough understanding of the current state of the art in device-free HGR on edge devices and in general. Finally, on the basis of present-day challenges and opportunities in this field, we outline which further research we suggest for HGR improvement. Our goal is to promote the development of efficient and accurate gesture recognition systems. Full article

Human activity recognition (HAR) using radar technology is becoming increasingly valuable for applications in areas such as smart security systems, healthcare monitoring, and interactive computing. This study investigates the integration of convolutional neural networks (CNNs) with conventional radar signal processing methods to improve the accuracy and efficiency of HAR. Three distinct, two-dimensional radar processing techniques, specifically range-fast Fourier transform (FFT)-based time-range maps, time-Doppler-based short-time Fourier transform (STFT) maps, and smoothed pseudo-Wigner–Ville distribution (SPWVD) maps, are evaluated in combination with four state-of-the-art CNN architectures: VGG-16, VGG-19, ResNet-50, and MobileNetV2. This study positions radar-generated maps as a form of visual data, bridging radar signal processing and image representation domains while ensuring privacy in sensitive applications. In total, twelve CNN and preprocessing configurations are analyzed, focusing on the trade-offs between preprocessing complexity and recognition accuracy, all of which are essential for real-time applications. Among these results, MobileNetV2, combined with STFT preprocessing, showed an ideal balance, achieving high computational efficiency and an accuracy rate of 96.30%, with a spectrogram generation time of 220 ms and an inference time of 2.57 ms per sample. The comprehensive evaluation underscores the importance of interpretable visual features for resource-constrained environments, expanding the applicability of radar-based HAR systems to domains such as augmented reality, autonomous systems, and edge computing. Full article

Millimeter-Wave Radar is one promising sensor to achieve robust perception against challenging observing conditions. In this paper, we propose a Radar Inertial Odometry (RIO) pipeline utilizing a long-range 4D millimeter-wave radar for autonomous vehicle navigation. Initially, we develop a perception frontend based on radar point cloud filtering and registration to estimate the relative transformations between frames reliably. Then an optimization-based backbone is formulated, which fuses IMU data, relative poses, and point cloud velocities from radar Doppler measurements. The proposed method is extensively tested in challenging on-road environments and in-the-air environments. The results indicate that the proposed RIO can provide a reliable localization function for mobile platforms, such as automotive vehicles and Unmanned Aerial Vehicles (UAVs), in various operation conditions. Full article

Network-group targets are a set of objectives that adhere to a shared communication protocol, perform common tasks, and exhibit relatively coordinated movements. Typically, network-group targets emit radar and communication signals. However, they often employ a silent mode to evade our perception. Despite this, they continue to exchange data through their communication networks. By intercepting the communication signals of these targets, this paper proposes a method for estimating the state and network topology of network-group targets based on random finite set (RFS) theory. This method is based on the modeling of network-group targets using a labeled multi-Bernoulli (LMB) RFS. In state estimation, the usual method involves extracting the localization parameters from the signals in the first step and use these parameters for target tracking in the second step. This study focused on estimating the kinematic states of network-group targets using communication signals containing delay and Doppler information received by multiple mobile sensor platforms. The proposed method considers the coherency between frequency-hopping pulses in the signals, resulting in an improved estimation performance. In addition, considering that network-group targets require network access for information exchange, graph theory concepts were utilized to estimate the network topology of network-group targets by address measurement. The simulation results validated the effectiveness of the proposed method and demonstrated its superior performance. Full article

Occupancy sensors are electronic devices used to detect the presence of people in monitored areas, and the output of these sensors can be used to optimize lighting control, heating and ventilation control, and real-estate utilization. Testing methods already exist for certain types of occupancy sensors (e.g., passive infrared) to evaluate their relative performance, allowing manufacturers to report coverage patterns for different types of motion. However, the existing published techniques are mostly tailored for passive-infrared sensors and therefore limited to evaluation of large motions, such as walking and hand movement. Here we define a characterization technique for a Doppler radar occupancy sensor based on detecting a small motion representing human breathing, using a well-defined readily reproducible target. The presented technique specifically provides a robust testing method for a single-channel continuous wave Doppler-radar based occupancy sensor, which has variation in sensitivity within each wavelength of range. By comparison with test data taken from a human subject, we demonstrate that the mobile target provides a reproducible alternative for a human target that better accounts for the impact of sensor placement. This characterization technique enables generation of coverage patterns for breathing motion for single-channel continuous wave Doppler radar-based occupancy sensors. Full article

Synthetic aperture radar (SAR) altimeters can achieve higher spatial resolution and signal-to-noise ratio (SNR) than conventional altimeters by Doppler beam sharpening or focused SAR imaging methods. To improve the estimation accuracy of waveform re-tracking, several average echo power models for SAR altimetry have been proposed in previous works. However, these models were mainly proposed for satellite altimeters and are not applicable to high-mobility platforms such as aircraft, unmanned aerial vehicles (UAVs), and missiles, which may have a large antenna mis-pointing angle and significant vertical movement. In this paper, we propose a novel semi-analytical waveform model and signal processing method for SAR altimeters with vertical movement and large antenna mis-pointing angles. A new semi-analytical expression that can be numerically computed for the flat pulse response (FSIR) is proposed. The 2D delay–Doppler map is then obtained by numerical computation of the convolution between the proposed analytical function, the probability density function, and the time/frequency point target response of the radar. A novel delay compensation method based on sinc interpolation for SAR altimeters with vertical movement is proposed to obtain the multilook echo, which can optimally handle non-integer delays and maintain signal frequency characteristics. In addition, a height estimation method based on least squares (LS) estimation is proposed. The LS estimator does not have an analytical solution, and requires iterative solving through gradient descent. We evaluate the performance of the proposed estimation strategy using simulated data for typical airborne scenarios. When the mis-pointing angles are within 10 degrees, the normalized quadratic error (NQE) of the proposed model is less than 10⁻¹⁰ and the RMSE of τ obtained by the re-tracking method fitted by the proposed model is less than 0.2 m, which indicates the high applicability of the model and accuracy of the re-tracking method. Full article

Deploying Passive Multistatic Radar (PMR) on mobile platforms provides covert and cost-effective monitoring over a large area, offering certain advantages in countermeasure. However, mobile PMR faces significant challenges, such as Doppler distortion and phase deviations. A multi-parameter space target detection method is proposed for mobile PMR to achieve target detection in three-dimensional environments. By estimating the Doppler Frequency Rate (DFR), applying bistatic range phase compensation, and implementing azimuth time integration, frame division, and data fusion, the detection accuracy and the Signal-to-Noise Ratio (SNR) are improved. Simulation results indicate that the proposed method significantly enhances the SNR and produces accurate detection results, demonstrating its efficacy. Full article

As one of the important vital features of the human body, the acquisition of a speech signal plays an important role in human–computer interaction. In this study, voice sounds are gathered and identified using Doppler radar. The skin on the neck vibrates when a person speaks, which causes the vocal cords to vibrate as well. The vibration signal received by the radar will produce a unique micro-Doppler signal according to words with different pronunciations. Following the conversion of these signals into micro-Doppler feature maps, these speech signal maps are categorized and identified. The speech recognition method used in this paper is on neural networks. CNN convolutional neural networks have a lower generalization and accuracy when there are insufficient training samples and sample extraction bias, and the training model is not suitable for use on mobile terminals. MobileViT is a lightweight transformers-based model that can be used for image classification tasks. MobileViT uses a lightweight attention mechanism to extract features with a faster inference speed and smaller model size while ensuring a higher accuracy. Our proposed method does not require large-scale data collection, which is beneficial for different users. In addition, the learning speed is relatively fast, with an accuracy of 99.5%. Full article

An aspect correlated with climate change is certainly represented by the alternation of severe floods and relevant drought periods. Moreover, there is evidence that changes in climate and land cover are inducing changes in stream channel cross-sections, altering local channel capacity. A direct consequence of a significant change in the local channel capacity is that the relationship between the amount of water flowing at a given point in a river or stream (usually at gauging stations) and the corresponding stage in that section, known as a stage–discharge relationship or rating curve, is changed. The key messages deriving from the present work are: (a) the more frequent and extreme the floods become, the more rapid the changes in the stream channel cross-section become, (b) from an operational point of view, the collection and processing of field measurements of the stage and corresponding discharge at a given section in order to quickly and frequently update the rating curve becomes a priority. It is, therefore, necessary to define a control system for acquiring hydrological data capable of keeping river levels and discharges under control to support flood early warnings and water management. The proposed stage–discharge management system is used by the Civil Protection Service of the Marche Region (east-central Italy) for the monitoring of river runoff in the regional watersheds. The Civil Protection Service staff performs stage–discharge field measurements using water level sensors and recorders (e.g., staff gauges, submersible pressure transducers, ultrasound and radar sensors) and a current meter, acoustic doppler velocimeter, acoustic doppler current profilers, portable mobile radar profiler and salt dilution method equipment, respectively. Power functions are fitted to the stage–discharge field data. Furthermore, extrapolation is performed to cover the full range of flow measurements; in general, extrapolation is not an easy task because of sharp changes in the stream cross-section geometry for very high or very low stages. In the present work, we also focused attention on the application problems that occur in practice and the need for frequent updating. Full article

The rescaling of rainfall requires measurements of rainfall rates over many dimensions. This paper develops one approach using 10 m vertical spatial observations of the Doppler spectra of falling rain every 10 s over intervals varying from 15 up to 41 min in two different locations and in two different years using two different micro-rain radars (MRR). The transformation of the temporal domain into spatial observations uses the Taylor “frozen” turbulence hypothesis to estimate an average advection speed over an entire observation interval. Thus, when no other advection estimates are possible, this paper offers a new approach for estimating the appropriate frozen turbulence advection speed by minimizing power spectral differences between the ensemble of purely spatial radial power spectra observed at all times in the vertical and those using the ensemble of temporal spectra at all heights to yield statistically reliable scaling relations. Thus, it is likely that MRR and other vertically pointing Doppler radars may often help to obviate the need for expensive and immobile large networks of instruments in order to determine such scaling relations but not the need of those radars for surveillance. Full article

The orthogonal frequency division multiplexing (OFDM) radar suffers from severe performance degradation in range-velocity estimation in high mobility scenarios. In this paper, a novel intercarrier interference (ICI)-free parameter estimation method for OFDM radar is proposed. By employing a scale discrete Fresnel transform (SDFnT), the OFDM radar signals are converted to the scale Fresnel domain, and the orthogonality of subcarriers can be recovered with the optimal scale factor. Furthermore, due to the compatibility of the SDFnT and the discrete Fourier Transform (DFT), the proposed method has low computational complexity and high feasibility for OFDM radar implementation. Simulation results show that the proposed SDFnT-based scheme effectively eliminates the ICI effect for single and multiple targets and achieves high accuracy delay-Doppler estimation for OFDM radar systems in circumstances of high velocity and low SNR with consistency and robustness. Full article

In order to exploit the advantages of CNN models in the detection of small floating targets on the sea surface, this paper proposes a new framework for encoding radar echo Doppler spectral sequences into images and explores two different ways of encoding time series: Gramian Angular Summation Field (GASF) and Gramian Angular Difference Field (GADF). To emphasize the importance of the location of texture information in the GAF-encoded map, this paper introduces the coordinate attention (CA) mechanism into the mobile inverted bottleneck convolution (MBConv) structure in EfficientNet and optimizes the model convergence by the adaptive AdamW optimization algorithm. Finally, the improved EfficientNet model is used to train and test on the constructed GADF and GASF datasets, respectively. The experimental results demonstrate the effectiveness of the proposed algorithm. The recognition accuracy of the improved EfficientNet model reaches 96.13% and 96.28% on the GADF and GASF datasets, respectively, which is 1.74% and 2.06% higher than that that of the pre-improved network model. The number of parameters of the improved EfficientNet model is 5.38 M, which is 0.09 M higher than that of the pre-improved network model. Compared with the classical image classification algorithm, the proposed algorithm achieves higher accuracy and maintains lighter computation. Full article