In this paper, we propose a novel protocol for discriminating multiple ballistic targets during the midcourse and terminal phases of their trajectories. The midcourse-phase protocol includes the following steps: 1) estimating the number of targets, 2) separating the trajectories, and 3) performing discrimination using either a convolutional neural network (CNN) or a nearest-neighbor classifier. The terminal-phase protocol performs discrimination using trajectory information or a CNN. Simulations that consider radar signals calculated using physical optics, indicate that the proposed protocol effectively discriminates between warheads and decoys in scenarios that involve multiple targets.

Compact radar sensors for Internet of Things applications can be used to analyze indoor human gait characteristics. Conventional human gait analysis methods typically involve generating two-dimensional (2D) high-resolution time-frequency images and employing image-processing techniques to estimate gait parameters of a walking human. However, these computations can be resource-intensive for the compact radar sensors. To address this problem, we propose a new scheme for estimating gait parameters. Our method has four significant contributions: 1) utilization of one-dimensional phase modulation in a radar echo for efficient gait-parameters estimation, as opposed to relying on 2D time-frequency images; 2) decomposition of micro phase modulations corresponding to the torso or pelvis and lower body parts (e.g., knee, tibia, and ankle) using dedicated filtering techniques to mitigate the interference between body components; 3) compensation for effects of nonlinear macro phase modulation caused by whole-body movements; and 4) robust estimation of gait parameters including time-varying radial velocity, gait rate, step length, and the height of the lower body. In experiments performed using a 5.8 GHz continuous-wave Doppler radar, we observed that the proposed scheme can perform efficient and robust gait parameter estimation of an indoor human walking.

In this paper, we propose a novel protocol for discriminating multiple ballistic targets during the midcourse and terminal phases of their trajectories. The midcourse-phase protocol includes the following steps: 1) estimating the number of targets, 2) separating the trajectories, and 3) performing discrimination using either a convolutional neural network (CNN) or a nearest-neighbor classifier. The terminal-phase protocol performs discrimination using trajectory information or a CNN. Simulations that consider radar signals calculated using physical optics, indicate that the proposed protocol effectively discriminates between warheads and decoys in scenarios that involve multiple targets.

Compact radar sensors for Internet of Things applications can be used to analyze indoor human gait characteristics. Conventional human gait analysis methods typically involve generating two-dimensional (2D) high-resolution time-frequency images and employing image-processing techniques to estimate gait parameters of a walking human. However, these computations can be resource-intensive for the compact radar sensors. To address this problem, we propose a new scheme for estimating gait parameters. Our method has four significant contributions: 1) utilization of one-dimensional phase modulation in a radar echo for efficient gait-parameters estimation, as opposed to relying on 2D time-frequency images; 2) decomposition of micro phase modulations corresponding to the torso or pelvis and lower body parts (e.g., knee, tibia, and ankle) using dedicated filtering techniques to mitigate the interference between body components; 3) compensation for effects of nonlinear macro phase modulation caused by whole-body movements; and 4) robust estimation of gait parameters including time-varying radial velocity, gait rate, step length, and the height of the lower body. In experiments performed using a 5.8 GHz continuous-wave Doppler radar, we observed that the proposed scheme can perform efficient and robust gait parameter estimation of an indoor human walking.

To detect a small target on the sea surface, we propose a detection algorithm that operates in three steps: 1) representation of a 3-D complex cube in the (radial range, pulse time, and scan time) domain by using stepped carrier frequency in fast scan mode; 2) intrascan integration to transform complex cube from the pulse time domain to the radial velocity domain by using Fourier transform; and 3) interscan integration to detect a small target by using a conventional optimal coherent detector with a Fourier relationship between stepped carrier frequency and initial radial range, and a discriminator between a true target and false targets. The proposed method improves the ability to detect a small target and reduces the sensitivity to sea clutter despite relatively low radial range resolution. In experiments, our proposed method could perform accurate and robust detection of a small target.

This paper proposes a recognition algorithm for the degree of rehabilitation of walking individuals using a Doppler radar sensor. First, the time-series macro velocities of the whole body were estimated by applying a moving average filter and differentiating the phase of the complex signal for the walking individual. Next, the time-series micro velocities of both legs were obtained by subtracting the time-series macro velocities of the entire body from the time-series velocities extracted by applying the conventional method and Kalman filter to the spectrogram. The step time and maximum velocities corresponding to each leg can be decomposed by estimating the local maximum values in the time-series microvelocities of both legs, resulting in new values to show the real-time degree of rehabilitation for walking individuals. In the experiments, we observed that the proposed method was capable of successfully recognizing both normal and abnormal walking movements.

ABSTRACT This letter presents a simple yet highly accurate markerless motion‐recognition method that uses radar. The proposed approach employs a freely available pose‐estimation tool (BlazePose) to extract 3D coordinates of a body while it is engaged in micro‐motions. These coordinates are used to generate basis signals, which are then combined with radar measurements to derive amplitude‐based features. Experimental results with a 77 GHz frequency‐modulated continuous wave radar show that the proposed method achieves classification accuracy that approaches 100%.

Abstract Respiratory and cardiac rates can be estimated by analyzing a spectrum of linearly mixed phase fluctuations in a radar echo of an individual. However, there are high-order harmonics caused by time-varying respiratory rate, and the interference effect of the respiratory rate and its harmonics makes it difficult to estimate the cardiac rate with relatively low energy in a spectrum. To solve this problem, we exploit the independent component analysis method with dual-band distributed continuous wave radar for effective decomposition of phase fluctuations corresponding to respiratory and cardiac rates. In simulations and experiments, the respiratory and cardiac rates were successfully estimated by the proposed decomposition method, compared with conventional methods.

Non-contact monitoring of oxygen saturation is required to diagnose obstructive sleep apnea, which leads to lethal sleep disorders in adults. This study proposes an estimation algorithm for oxygen saturation using a noncontact radar sensor. First, the algorithm estimates the displacement of the chest, respiratory rate, and cardiac rate using the phase fluctuations of complex signals received from an individual. Next, the pulmonary ventilation, tidal volume, and minute ventilation were estimated using a step-by-step process, leading to the prediction of pressures corresponding to carbon dioxide and oxygen. Finally, the oxygen saturation was calculated by a second-order transfer function using the predicted pressures. In particular, the proposed method can robustly estimate oxygen saturation despite clutter and unwanted body movements because of the use of phase fluctuations preprocessed by several filtering methods. Through simulations and experiments, we observed that the proposed method using a Doppler radar can robustly estimate the oxygen saturation of an individual with unwanted body movements, unlike conventional methods.

본 논문에서는 병진 및 회전에 대한 불변성을 갖는 역합성 개구면 레이더(ISAR, Inverse Synthetic Aperture Radar) 영상 식별을 위한 효과적인 방법을 제안한다. 제안된 방법은 학습 영상과 시험 영상 간의 상대 회전각을 추정하기 위해 ISAR 영상에 대한 이차원 푸리에 변환을 수행한 후, 해당 스펙트럼을 극좌표계로 사상한다. 학습 영상과 시험 영상의 사상 영상 간 상관계수가 최대가 되는 각도를 회전각으로 설정한 후, 시험 영상의 2D 푸리에 스펙트럼을 회전각 만큼 회전시키고, 이를 이차원 역푸리에 변환하여 회전에 대한 불변성을 확보한다. 영상 식별은 이차원 주성분 분석과 2개 은닉층의 간단한 신경망 분류기를 통해 수행된다. 제안된 방법은 적은 수의 학습 데이터와 낮은 신호대잡음비 환경에서도 높은 식별 정확도를 나타내며, 기존 방법에 비해 영상의 흐림 현상에 덜 민감한 특성을 보인다.

Sea skimming and high-maneuverability anti-ship cruise missiles have been designed to evade the surveillance and tracking of maritime multi-functional radar (MFR), resulting in difficulties in radar target tracking. Sea skimming and high maneuvering cause two problems: distortion of the estimated elevation angles by multi-path echoes and overload of radar beam resources by precise tracking tasks. Recently, a dual-band MFR has been developed for state-of-the-art battleships. In this study, a dual-band MFR system is used to improve the tracking accuracy of sea skimming and high-maneuvering targets. In particular, the fusion and allocation of dual-band MFR echoes to address the problems caused by multi-path echoes and heavy precision tracking tasks can lead to improved tracking accuracy. The simulation results revealed that our proposed method is capable of performing successful tracking of sea skimming and high-maneuvering targets in a dual-band MFR system.