Compressive sensing (CS) technology has proven highly effective in rapid data acquisition and super-resolution target imaging for through-the-wall radar imaging (TWRI) applications. However, most existing CS-based TWRI techniques focus primarily on high-quality imaging of behind-wall targets, often neglecting the reconstruction of building layouts, which is essential for determining the relative positions of targets in unknown environments. To address this limitation, a structural-prior-guided CS framework is proposed for the joint reconstruction of building layouts and behind-wall targets. Specifically, first, distinct imaging models are developed for layouts and targets, accounting for their unique structural properties: layouts, referring to wall structures, typically manifest as extended, piecewise-continuous line-like structures, whereas targets manifest as compact, point-like structures. Building on these models, a unified constrained optimization problem is formulated by integrating (i) the strong inter-channel correlation of layout echoes, enforced via a low-rank regularization on the layout component, and (ii) structured sparsity priors tailored to both the layout and target images. Then, the resulting composite problem is efficiently solved using proximal gradient algorithm, yielding simultaneous reconstruction of the unknown building layouts and behind-wall targets. Finally, simulations and experimental results demonstrate the effectiveness of the proposed algorithm.

The sparse array configuration of multi-input multi-output imaging radar leads to high grating lobes problem in the imaging process, which significantly degrades final image quality. Although the traditional Phase Coherence Factor can partially mitigate these grating lobes, it suffers from limitations such as attenuation of the main lobe energy. To overcome these drawbacks, this paper proposes a novel grating lobes suppression method based on phase-coherence-guided adaptive threshold classification. This method first adaptively determines a classification threshold by analyzing the phase coherence features of the target main lobe. Using this threshold, all the grids in the radar image are classified into two categories and distinct schemes are applied to compute their respective weighting factors. Finally, grating lobes in the image are suppressed by weighting the original radar image. Numerical simulation and field experiment both confirm the effectiveness of the proposed method, which achieves a higher peak sidelobe ratio than conventional methods, demonstrating promising practical value.

Millimeter-wave (mmWave) radar and camera are two of the most critical sensors in modern intelligent transportation systems (ITS), enabling complex fusion-based perception tasks through collaborative working. Achieving high-quality data fusion in ITS requires precise extrinsic parameters (EPs), which define the relative spatial relationship between sensors. However, in real traffic environments, manual measurement of EPs between sensors is labor-intensive and limited in accuracy. To overcome these challenges, this paper proposes an automatic extrinsic calibration method for mmWave radar and camera in traffic environments, requiring only time-synchronized sensor data. First, we develop a novel calibration model with eight parameters, the radar-camera-ground (RCG) model, which describes the spatial relationships between the sensors and the ground. Then, a calibration method named Gaussian Modeling Linear Optical Projection (GLP) is proposed. Specifically, an image instance segmentation model is applied to detect targets from images. Simultaneously, 2D information of targets detected by the radar is extended into 3D space according to the RCG Model. Next, vehicle targets detected by radar and camera are transformed into corresponding 3D and 2D Gaussian models, respectively, leveraging their positional, velocity, and shape features. Afterward, a mapping relationship between the 3D and 2D Gaussian models is established through a linearized optical projection function. Finally, the optimal EPs are estimated via the global optimization algorithm, minimizing the designed calibration loss function based on Bhattacharyya distance. Experimental results on practical traffic scenario data demonstrate that the proposed method outperforms existing approaches in average calibration accuracy and robustness, validating its reliability and superiority.

In this paper, a vision-assisted multipath recognition and suppression method is proposed for the problem of millimeter wave (mmWave) radar producing false targets under the influence of multipath interference. First, object detection is performed on the image and rectangular clustering is performed on the mmWave radar point cloud to complete the data preprocessing. Subsequently, nearest-neighbor frame matching and direct linear transform (DLT) algorithms are used to achieve spatio-temporal calibration of the two sensors. An axial adaptive cost-normalized matching algorithm is then proposed to associate targets from the two sensors, thereby establishing target association pairs. Finally, multipath ghosts in mmWave radar are recognized and suppressed based on the target association results. Experimental results show that the proposed method efficiently recognizes and suppresses multipath ghosts in traffic scenarios.

Non-line-of-sight (NLOS) detection is crucial in applications such as autonomous driving and surveillance. This paper proposes a bistatic multiple-input multiple-output (MIMO) radar-based joint estimation algorithm to localize diffraction corners and estimate NLOS targets. By leveraging the direction of departure (DoD) and direction of arrival (DoA) of diffraction signal, the algorithm first estimates corner position. Furthermore, target motion state is estimated based on the estimated corner and Doppler information. Electromagnetic simulations confirm the accuracy and robustness of the proposed method under various noise conditions.

This study introduces an improved signal de-composition methodology for non-contact heartbeat estimation using millimeter-wave (mmWave) radar. With the increasing demand for non-invasive and continuous monitoring of vital signs, mmWave radar technology has become a promising alternative to traditional contact-based methods, such as electrocardiogram (ECG), due to its high sensitivity, robust penetration, and adaptability to diverse environments. Specifically, we first analyze the signal of the mmWave radar system to derive a model of cardiac signal extraction based on radar echo signal. Variational Mode Decomposition (VMD) integrated with a Newton-Raphson-based optimizer (NRBO) algorithm is then utilized for the accurate reconstruction of the cardiac mechanic signal (CMS). The VMD method decomposes the signal into its intrinsic mode functions (IMFs), while the NRBO dynamically optimizes the decomposition parameters, including the penalty factor (α), to enhance the precision of the heartbeat estimation. The effectiveness and robustness of the proposed model is validated through a 18 subjects experiment dataset, and the model shows significant improvements over three baselines in terms of accuracy and reliability of heartbeat detection.

In the realm of human activity recognition (HAR) based on radar, the prevailing methods have been characterized by excessive complexity and a singular focus on a specific motion direction, posing challenges for practical deployment. This article introduces a lightweight parameter estimation and deep learning hybrid extraction network (PDHE-Net) for achieving multidirectional HAR based on mmWave radar. The network consists of a lightweight deep learning feature extraction (LDE) module, a parameter estimation module, and a classification module. Specifically, the LDE module consists of multiple layers of group convolution and Ghost module, aimed at extracting deep features from the time-Doppler (TD) maps. Parameter estimation module is employed to capture direction-independent features from the TD map. Ultimately, multidimensional features extracted by the LDE module and parameter estimation module are classified to realize multidirectional HAR. To verify the performance of the proposed method, experimental data was collected, comprising six categories of activities wherein targets moved in multidirections. The experimental results demonstrate that the proposed PDHE-Net achieved a recognition accuracy of 96.67% on the dataset, outperforming the state-of-the-art methods by 2.92%, while significantly reducing complexity.

Radar-based human activity recognition (HAR) holds significant application value in fields such as medical rehabilitation and security monitoring. However, existing HAR methods primarily address line-of-sight (LOS) and non-line-of-sight through-wall (NLOS-TW) scenarios, neglecting consideration for non-line-of-sight corner (NLOS-C) scenario within urban architecture. In NLOS-C scenario, electromagnetic waves illuminate the target through multiple paths, resulting in considerable variations in range-time map, causing performance degradation or even failure of existing methods. Moreover, multipath propagation enables a single-node radar to function as a multi-perspective multi-node radar system, providing more comprehensive information for human activity. Therefore, considering the complementary interpretations of multipath and the distinctive features observed in NLOS-C range-time map, this paper proposes a HAR method for NLOS-C scenario based on multipath fusion. Firstly, considering the broad distribution and large span characteristics of behavior features in the range dimension caused by multipath effect, we design a multipath information fusion module based on dilated convolution to effectively integrate and interact the multipath information. Additionally, to address the diverse feature scales caused by variable widths and blurred boundaries of each path, we incorporate multi-scale unit into the deep feature extraction module to enhance the capability of autonomously adjusting receptive field. Finally, multipath interaction information is fused with depth features for recognition. Experimental results validate the effectiveness of the proposed method in NLOS-C scenario. The codes are available at https://github.com/tlz1111/Multipath-Fusion-Network.

This paper considers the problem of line of sight (LOS) and non-line-of-sight (NLOS) targets localization in an L-shaped corner. Specifically, first, the propagation paths of the electromagnetic waves in an L-shaped corner are analyzed by the ray tracing technology. Then, a sparsity-based multipath model is formulated by regularizing the LOS target with sparsity norm and the NLOS targets with jointly sparse across different paths, respectively. After that, a proximal gradient-based iterative algorithm is proposed to tackle this optimization problem. Finally, the feasibility of simultaneously localizing NLOS and LOS targets is demonstrated through simulations.

Methods by radar spectrograms are critical technologies in the field of human activity recognition. However, most methods are limited by the single domain feature expression and the use of a single time-frequency (TF) analysis method in the TF domain. In this paper, we propose a multi-domain fusion human activity recognition network based on multi-resolution TF spectrograms and range spectrograms for human activity recognition. The attention mechanism is utilized to extract spectrograms with different features in the range and TF domain. In the TF domain, the intrinsically relevant features of the three types of spectrograms are extracted using the 3-D convolutional neural network (3DCNN). Then the bilinear pooling is used to fuse the features of two different domains so that the features of the activity are comprehensively characterized. Finally the fused multi-domain features are extracted using the 2-D convolutional neural network (2DCNN). The results of the ablation experiments verify the validity of the proposed model.

This paper considers the building layout reconstruction (BLR) problem based on compressive sensing (CS) framework in the impulsive noise environment. Specifically, first, a CS-based imaging model considering the extended characteristics of walls is established. To realize effective BLR under impulsive noise, we formulate an optimization problem integrated maximum complex correntropy criterion (MCCC) and sparsity constraint. Then, a proximal-gradient-based iterative algorithm is introduced to solve the optimization problem. Finally, simulations under Gaussian noise and impulsive noise validate the effectiveness of the proposed method.

The closure phase quantifies phase inconsistency caused by scattering phenomena such as vegetation, roughness, and soil moisture variations. This study explores the scattering characteristics of the closure phase. We use the singular value decomposition (SVD) method to separate the scattering phase from the closure phase. Then, we investigate the scattering properties associated with the zero and non-zero closure phase through H/α polarization decomposition in cropland. The results show that the zero closure phase area primarily corresponds to surface scattering, with a low entropy area (64.28%). However, the non-zero closure phase area is linked to medium-to-high entropy levels, implying a more complex scattering process (54.67%). This study clarifies the physical scattering properties behind closure phase values in cropland and provides insight for applications.

In this paper, we consider the problem of multipath suppression for through-the-wall radar imaging (TWRI). Exploiting the spatial diversity characteristic of multiple-input-multiple-output (MIMO) radar, we present a novel framework integrated compressive sensing and spatial filtering to formulate a ghost-free image with high-resolution. Specifically, an array signal processing model that accounts for the direction of departure (DoD), direction of arrival (DoA) and time of arrival (ToA) is established, where the signal with unequal DoD and DoA is regarded as the multipath. For the purpose of computational efficiency, we present a two-step method to solve the proposed large-scale problem, where the alternating direction method of multiplier (ADMM) and Capon beamforming are employed. Finally, the performance of the proposed method is verified via simulations.

Gesture recognition in few-shot scenarios presents a significant challenge due to the scarcity of labeled data. In this work, we propose a novel end-to-end framework tailored for few-shot gesture recognition, addressing the limitations of current methods. A novel feature map generating method is proposed to leverage a greater number of dimensions in capturing gesture feature information and simplify the structure of network. Our approach also maximizes the utility of a limited set of real training samples by generating new virtual samples in two domains based on data augmentation, and employs a feature fusion strategy for comprehensive gesture characteristic extraction by using both Convolutional Neural Network (CNN) and Histogram of Oriented Gradients (HOG) to extract features. Extensive experimental results validate the efficacy of our proposed method, achieving a final accuracy of 85.26%, exhibiting a remarkable 35% improvement over the baseline, thereby confirming the effectiveness of our work in the challenging few-shot gesture recognition task.

As a 3-D point cloud has the ability to present the contour of an object clearly, it provides more spatial information for person identification (PI) task. Aiming at the improvements on the quality of point cloud and distribution of features, an innovative treatment method for point cloud and a novel network structure are investigated in this article. First, spatiotemporal feature of point cloud is enhanced by implementing dual-stage density-based spatial clustering of applications with noise (DST-DBSCAN) method, which can filter most invalid points and decrease the sparsity of point cloud. After that, the optimized point cloud is input into neural network, which contains three parts for feature extraction, classification and feature optimization. Specifically, PointNet++ is adopted to extract features and realize PI recognition. In addition, an adversarial network is designed for optimizing feature distribution of point clouds by encouraging the feature extractor of PointNet++ to generate features of the same person as similar as possible. Experimental results demonstrate that the proposed method can improve the accuracy by 3.77% than original PointNet++ network with raw data.

This letter considers the problem of target detection and tracking with millimeter-wave (mm-wave) frequency-modulated continuous-wave (FMCW) radar and proposes a tracking method via learning the displacement of the target in successive frames. First, a convolutional neural network (CNN), namely, detection and displacement network (DDNet), is trained to predict the target positions at current frame and the displacement relative to last frame simultaneously with the range-azimuth (RA) spectrum in two adjacent frames. Then, the detections are associated through the predicted displacement to construct the trajectories. Finally, the effectiveness of the proposed method is validated on both simulated and real-world datasets.

In the field of through-the-wall radar imaging , existing compressive sensing (CS) methods mainly concentrate on deriving indoor targets image while overlooking the reconstruction of building layout image. In this paper, we focus on the problem of utilizing CS for building layout reconstruction (BLR) in wall reverberation environment. Specifically, first, by incorporating the characteristics of building layout, an extended target CS imaging model in wall reverberation environment is established. Then, an extended-target-based group block CS (ET-GBCS) algorithm based on the alternating direction multiplier method is proposed to accurately reconstruct the building layout. After obtaining the reconstructed result of each view, the total variation minimization method is used to process the multi-view fusion result for building layout edge preservation and image noise removal. Finally, the effectiveness of the proposed algorithm is verified by electromagnetic simulations.

The pursuit of high-quality building layout images is a key objective in radio tomographic imaging (RTI) as it provides essential information for precise indoor target localization. This study addresses the challenge of tomographic imaging for three-dimensional (3D) building layout, introducing a tensor-based enhancement imaging method. Specifically, first, the linear tomographic model is built by considering the relationship between the time delay of the transmissive signal and unknown region. By solving the tomographic model, the initial spatial map can be derived, and it is characterized as a three-order tensor, encapsulating the spatial attributes of the building. In the proposed enhanced imaging method, it leverages the spatial correlations, smoothness, and adaptive group sparsity properties inherent in 3D building layouts, and embeds those prior knowledge into the tensor-based optimization framework, which not only enhances reconstruction accuracy but also suppresses the striping artifacts. Numerical simulations and experiments are conducted to validate the proposed algorithm, with comparisons made against state-of-the-art methods. The results consistently demonstrate a substantial improvement in the quality of building layout image, which underscores its high potential and applicability within the field of radio tomography.

Building layout tomographic imaging (BLTI) is a nascent technology that effectively addresses issues, such as wall cavities and offsets, encountered in traditional synthetic aperture radar (SAR) imaging. Nevertheless, the current single-input single-output (SISO) imaging system is susceptible to multipath signals (MPSs) interference, leading to issues, such as artifacts and distortions in the imaging results. In this article, a tomographic scanning system based on multiple-input multiple-output (MIMO) ultrawideband (UWB) radar is designed, which can accurately acquire the delay of direct-path signals (DPSs) delay by the similarities of DPS across different channels. Next, a linear tomographic projection model is established according to the relationship between DPS delay and the scene to be imaged. Furthermore, a joint sparse imaging algorithm with adaptive projection matrix modification (JS-APMM) is proposed. It takes into consideration the prior information of building layout and material to improve imaging quality under limited data. Finally, simulation and experimental results demonstrate the superiority of the proposed method over the single-input single-output (SISO) imaging system and state-of-the-art compressed sensing algorithms.

Building layout reconstruction (BLR) is a prominent research topic in the field of through-the-wall radar (TWR) and wireless perception. Inspired by computed tomography (CT), the transmit–receive separated dual-bistatic radar system utilizes electromagnetic (EM) wave transmission signals to perform BLR. However, existing researches significantly rely on signal frequency bandwidth resources. Moreover, state-of-the-art researches only estimate the time delay (TD) information, thereby posing challenges in precisely discriminating between the direct path (DP) and multipaths. This article refines the sparse signal reconstruction-based angle of arrival (AOA) and TD super-resolution estimate algorithms under the condition of restricted broadband array signals. In accordance with this, the present study proposes a DP-identifying criterion with the assistance of AOA and obtains high-accuracy DPTD estimation. Furthermore, with the high-accuracy DPTD estimation, this article proposes a common material permittivities-based iterative multimaterial BLR algorithm. The final numerical simulations and EM simulations verify the effectiveness of the proposed super-resolution algorithm and the improvement of multimaterial BLR.

In the field of ultra-wideband (UWB) multi-input multi-output (MIMO) imaging, it is difficult to recognize targets, especially weak targets, in the high grating lobes environment. To solve this problem, in this paper, an adaptive sub-band sub-aperture coherence factor (ASBSA-CF) algorithm is proposed based on sub-band sub-aperture (SBSA) image data. First, we establish the objective function by using the maximum inter-class variance (OTSU) method. Then, particle swarm optimization (PSO) algorithm is applied to adaptively choose the number of sub-bands and sub-apertures of SBSA image data. Finally, simulation verifies that ASBSA-CF can optimally achieve the compromise between weak targets and grating lobes suppression.

Ghost targets resulting from the reflection of electromagnetic waves by stationary objects are a long-standing issue. Currently, the majority of techniques for addressing multipath ghost targets involve either prior awareness of the environment or cooperation between multiple radars, which has its drawbacks. To address the problem with a single radar, this paper proposes a multipath ghost recognition method based on the difference between direction of departure (DOD) and direction of arrival (DOA) for sparse multiple input multiple output (MIMO) radar that does not require prior building information. This method first achieves DOD and DOA estimation of the signal through minimum variance distortionless response (MVDR). Then, a grating lobe suppression method for sparse MIMO radar in multipath environments is proposed based on the distribution characteristics of DOD-DOA spectral peaks of multipath signals. In that case, the recognition of1-ordermultipath is completed by using the features of two types1-ordermultipath DOD and DOA being unequal. After that, the recognition accuracy of1-orderand non-combined multipath signals is improved by accumulating multiple frames. On this basis, the linear relationship among the distances corresponding to0-ordermultipath,1-ordermultipath and2-ordermultipath, is utilized to complete the recognition of2-ordermultipath signal. Finally, several experiments are carried out to verify the effectiveness of the proposed method.

In this paper, we consider a high-accuracy and low-complexity method for recognizing human activities behind wall. As the amount of information conveyed by data representation directly affects the recognition accuracy of the network, we construct the three-dimensional (3D) complex-valued feature for human activity recognition (HAR). In light of high network complexity introduced by 3D complex-valued data, we devise a low time and space complexity network named Convolutional Gated Recurrent Unit based on region-vectorization (RV-ConvGRU). Keystone Transform is utilized to process the radar echo and generate 3D complex-valued Range-Time-Doppler (RTD) data first, which provides high-frequency resolution and abundant feature information. Then, the real and imaginary parts of the complex-valued RTD are separately fed into a feature extraction module to comprehensively extract their respective features. Specifically, the real or imaginary part of the RTD is divided into multiple regions, which are then converted into regional vectors and reordered as channels to reduce the time and space complexity of the subsequent network. The reconfigured features are then input into the Convolutional Gated Recurrent Unit (ConvGRU) to extract global and temporal features, with the channel attention mechanism for feature selecting. The features of the real and imaginary parts are fused and then classified by the classifier finally. The experiments verify that the proposed method is effective, achieving the highest recognition accuracy of 99.23% with an input sequence of 1.44 seconds.

Localization of non-line-of-sight (NLOS) targets in the complex urban environment have attracted significant attention in recent years. However, the requirement for precise prior information about the environment is idealistic. It is challenging to know the environmental information in the blind area of vision in advance of practical applications. This paper proposes a joint estimation algorithm for building layout and target position in the L-shaped scene without any prior information. Specifically, a round-trip multipath propagation model is first developed for the cases of diffraction and multiple reflections. Then, the received echo signal is preprocessed with moving target identification (MTI), back-projection (BP) imaging, and image segmentation. In addition, the target points, which are screened by geometric association, are further matched and estimated by the multipath ghost’s hypothesis method, thus realizing the joint perceptual estimation of the building layout and the target position. Finally, electromagnetic (EM) simulations and experimental measurements are used to validate the effectiveness of the proposed algorithm.

Human motion recognition based on ultra-wideband through-the-wall radar (UWB TWR) (a radar whose fractional bandwidth of the radar transmitted signal is bigger than 0.25) is faced with the problems of too few samples and the limitation of perspective. In this letter, we propose a multiradar cooperative human motion recognition model based on transfer learning and ensemble learning. Specifically, a ResNeXt network model based on transfer learning is first proposed to deal with the problem of too few samples. The model is pretrained on the public ImageNet database, and then it is transferred to the task of human motion recognition based on multiradar. Compared with a typical convolutional neural network from scratch, the ResNeXt network model based on transfer learning requires shorter epochs and achieves higher accuracy. Then, to solve the problem of model accuracy decline caused by the limitation of perspective, a multiradar human motion recognition model based on ensemble learning is proposed. Experimental results show that compared with the fusion model based on single-view radar, the recognition accuracy of network based on ensemble learning can be higher.

Through-wall human motion recognition is suffered from the problems of too few samples and too large model parameters. In this letter, we propose a multi-domain fusion through-the-wall radar (TWR) human motion recognition model based on lightweight network and transfer learning. Specifically, in order to make full use of the target information, a multiple parallel feature pyramid network (FPN) is first proposed to extract the detailed feature information from the time–frequency map and range profile. After that, a lightweight network based on the MobileNetV3 network and transfer learning is proposed. The MobileNetV3 model is pre-trained on the public ImageNet database. To ensure the performance of transfer learning, a heterogeneous migration learning algorithm is used to cross-domain transform the obtained time–frequency map and range profile. Experimental results show that the proposed model has a better performance in accuracy, model size, training time, and robustness compared with the existing methods. It also has the potential to embed portable radar, which has important research value for the application of radar in real life.

This letter studies the problem of ultrawideband (UWB) tomographic imaging for unknown building layouts where the multipath-rich condition is considered. Specifically, first, the multiple propagation paths of the UWB signal are analyzed, and a delay estimation algorithm is proposed to estimate the direct path (DP) delay from the multipath signal. Then, a tomographic projection model is established by mapping the relationship between the delay of the DP and the relative permittivity of the unknown layout. Besides, a modified total variation method is developed to reconstruct the building layout, which shows significant performance in preserving the edges of the structure. Finally, the effectiveness of the proposed algorithm is verified using both simulated and real data.

This letter considers the problem of human activity recognition (HAR) behind the walls using ultra-wideband (UWB) radar. The graph convolutional network (GCN)-enhanced multi-domain fusion network (GMFN) is proposed to improve the recognition performance utilizing the complementarity of the multi-domain features. Specifically, firstly, a multi-branch convolutional neural network (CNN) is proposed to extract the multi-domain features from the range, time-frequency, and range-Doppler domain. Then the multi-domain features are constructed as a graph, and the GCN is employed to fuse the multi-domain features on the graph. Finally, HAR is implemented in the form of the graph classification. The experimental results on the real data show that the proposed GMFN achieves better performance than the state-of-the-art multi-domain fusion HAR methods.

Building layout reconstruction (BLR) is an important topic in the field of through-the-wall imaging. Traditionally, reflective signals are commonly used to generate an accurate building map. However, due to the inherent features of electromagnetic waves, the reconstructed walls will inevitably suffer from problems such as deviation and cavities. Alternatively, as an extension of computed tomography, the transmissive signals can be exploited for BLR with high efficiency, but its performance degrades seriously when the sampling views are sparse. In this paper, to fully combine the superiority of the two types of implementations of BLR, we proposed a hybrid imaging framework by jointly exploiting the reflective and transmissive signals to retrieve the unknown layout. Specifically, first, the time delay of the transmissive signal will be estimated and used to reconstruct the spatial tomographic map. Then, a series of reflected echoes sampled by different routes will be compensated iteratively and used to generate the back-projection image. Finally, we fuse the two images generated by different types of signals using the feature-level detector. Both simulated and experimental results reveal that the proposed imaging framework can yield better performance compared with the image derived from single-type signals.

Non-line-of-sight (NLOS) detection is an enduring topic as it provides a powerful tool to monitor visually blocked areas. Currently, the NLOS detection requires precise prior knowledge of building layout, which limits its further applications in practice. In this paper, we consider the problem of joint estimation of building layout and target location in the NLOS scenario by exploiting multipath returns. Specifically, first, the building layout is simplified into combined linear equations with unknown parameters. In this way, we establish a parametrized multipath propagation model in the multiple targets NLOS scenario for the multiple-input-multiple-output (MIMO) radar, which is used in the image reconstruction and layout estimation problem. Then, a shape-remodeling group sparse constraint algorithm is proposed and combined with the particle swarm optimization method to simultaneously reconstruct the unknown layout and targets. Compared to the conventional compressed sensing-based methods, the proposed method integrates the basic structural characteristics and sparsity prior of the NLOS image to improve the stability of the solution. Finally, the performance of the proposed method is verified with numerical and experimental results.

Recently, wireless communication signals are not only used for information transmission but other possibilities are being developed, such as human body state estimation, wireless sensing, etc. In this article, the radio tomographic reconstruction of building layout and interior object is considered only utilizing the received signal strength indication (RSSI) of commercial WiFi signals. Specifically, first, an electromagnetic transmission energy attenuation model is introduced, which considers the case for scanning the entire unknown area using two wireless nodes. Then, we propose a novel algebraic iterative reconstruction method, which takes into account the constraint of prior knowledge and designs a more reasonable initial iteration factor. Finally, a variety of scenarios with different structures is built, and the reconstruction of the multimedia scenario is achieved for the first time with RSSI information. The simulation and experimental results show that the proposed method has better performance than current state-of-the-art algorithms in both the single-medium scenario and multimedia scenario.

Shape variance of target images, image overlapping for adjacent targets, and weak scattering target detection are critical challenges of human target detection for through-the-wall radar imaging. In this letter, an adaptive target detection method is proposed based on fully convolutional network (FCN). The downsampling-upsampling structure is employed to extract multiscale features. The attention mechanism is integrated with the FCN for weak scattering target detection. Exploiting both the intensity and geometrical features of the target image, the proposed algorithm could overcome the abovementioned challenges and achieve better detection performance compared with the state-of-the-art methods. The proposed algorithm is evaluated via simulation and experimental tests.

This paper proposed a background-noise self-adaptive voice activity detection (VAD) accelerator using SNR prediction based precision dynamic reconfigurable approximate computing. To improve the energy efficiency while maintaining high recognition accuracy for different background noises, two optimization techniques are proposed. Firstly, we proposed a SNR prediction module to analyze and pre-classify the back-ground noise into different levels, and a binarized weight network (BWN) accelerator with reconfigurable data bit width to implement the feature classification of VAD. Then, we proposed an approximate computing architecture with precision self-adaptive approximate addition unit to further reduce the energy consumption of BWN accelerator. Evaluated under 28nm process technology, this work can achieve high recognition accuracy (speech/none-speech hit rate: 95%/92% @10dB, 90%/87% @5dB, and 85%/80% @-5dB) under different background noise (SNR-5dB) with a low power consumption of 2 ~ 8uW.

In this letter, we consider the problem of human target detecting and tracking, exploiting small-aperture through-wall imaging radar. We build a novel target model considering both the statistical and the geometrical information of the target image. A scale-adaptive target tracking method is proposed to track the scale and orientation variant human targets based on the mean-shift tracking framework, where the image moments are exploited to estimate the scale and orientation of the target image dynamically. Finally, the proposed algorithm is evaluated by simulations and experimental results.

This paper solves the through-wall imaging (TWI) problem with a narrow-band system, and proposes an adaptive TWI method based on data fusion of multiple scan paths. First, we use a Wentzel-Kramers-Brillouin-based (WKB-based) approximation to model the interaction of the transmitted wave with the unknown area. Then we use Radon inverse transform to reconstruct the image from the received signal strength data of different paths. Furthermore, we evaluate the impact of scan paths on imaging. Finally, finite-difference time-domain (FDTD) simulation results demonstrate the validity of proposed method.

This paper deals with the tracking problems for multiple human targets hidden behind the wall using through-wall imaging radar (TWIR). We propose a robust tracking algorithm in image domain, combining mean-shift algorithm with Kalman filter. Comparing with the traditional mean-shift algorithm, the proposed algorithm has a greater performance in multiple human targets tracking, especially considering the case of the temporary loss of target. Real data validates the robustness of the proposed algorithm.

This paper considers the location problem for Moving targets behind a corner. Exploiting multi-path and the algorithm based on phase comparison among the multiple channels can obtain the position of the target behind a corner. To localize the moving target, a scanning radar system with multiple channels is suggested. The false target range can be achieved by the fast Fourier transform(FFT) technique. In addition, the false target azimuth is derived via exploiting the phase differences between the return signals among the multiple channels. Due to false targets and real targets are geometric symmetry, true targets can be localized by the radar system. Finally the experiment results validate this method, and demonstrate the effectiveness.

In this letter, we consider the problem of multipath ghost suppression for through-the-wall imaging radar. Exploiting the fact that these locations of the multipath ghosts depend on while the target location is independent of the array configuration, we present a novel framework via array rotating to eliminate the multipath ghosts. Specifically, we first rotate the array with multiple different array rotation angles. Then, multiple images are derived using back-projection imaging algorithm. Finally, the incoherent arithmetic fusion method is applied to yield a ghost-free image. The proposed approach has two advantages for multipath ghost suppression. First is the simplicity of the operation, and the second is that it will not be affected by the incorrect wall parameters. Ghost suppression performance of the proposed approach is evaluated via numerical simulations.