8 April 2024 in IEEE Antennas and Wireless Propagation Letters
Non-contact detection of vital signs, such as respiration and heartbeat, can be implemented with high accuracy using commercial radar sensors. However, the task is more challenging for non-line-of-sight (NLOS) scenarios, wherein the direct path between the radar and target is blocked. In this letter, we propose a method for detecting the vital signs of an NLOS target using the multipath propagation characteristics of the ultra-wideband (UWB) radar. The UWB radar is suitable for the NLOS target detection because the multipath-rich channel information can be estimated using an extremely narrow pulse with a wide bandwidth. By analyzing the multiple propagation paths of the NLOS scenario, we demonstrate that the second-order path is the most appropriate for vital sign detection. The experimental results showed that the proposed NLOS vital sign detection method can estimate respiration and heartbeat frequencies within the frequency resolution.
19 November 2024 in IEEE Transactions on Instrumentation and Measurement
Indoor human detection is a key prerequisite for realizing smart home systems. However, radar sensors for human detection experience severe performance degradation due to multipaths of indoor environments. Multipath signals generate ghost targets that highly deteriorate radar system stability. In this study, we analyze the limitations of the conventional multiple-input and multiple-output (MIMO) beamforming method using the virtual antenna array by considering two failure cases with respect to multipath suppression. To overcome such limitations of the conventional beamforming, a path matching algorithm is proposed based on the MIMO frequency-modulated continuous-wave (FMCW) radar. The path matching algorithm leverages the angle properties of signal paths to identify direct paths and the first-order multipath pairs. It utilizes 2-D angle analysis by considering the angle of departure (AoD) and angle of arrival (AoA) separately and also addressing the angle ambiguity problem caused by large antenna spacing. Experimental results verify that the proposed algorithm has the superior detection performance, in terms of both the number of ghost targets and missed targets, compared with the conventional beamforming method using the virtual antenna array.
18 October 2024 in IEEE Transactions on Instrumentation and Measurement
In this paper, we propose an eye-gaze tracking method based on head orientation estimation that uses a single 60 GHz frequency-modulated continuous-wave (FMCW) radar sensor. The FMCW radar data are acquired for cases, in which the radar is illuminating the front or side of the face. Because the variation in facial muscles caused by eye blinking is more pronounced when a human gazes at the radar, the orientation of the human head can be estimated by analyzing the received radar signal. First, an approximate range-angle map is generated to identify whether a human exists. When a human is detected, the received signal is projected onto the lower subspace of interest. Subsequently, a super-resolution algorithm is applied to the projected signal to obtain a precise target spectrum. The accumulated spectrogram is used as input to MobileNet to classify radar signal images corresponding to different orientations of the human head. The classification results show that the proposed method can identify the orientation of a human head with an accuracy exceeding 90%.
1 January 2026 in IEEE Access
30 January 2026 in IEEE Sensors Journal
Conventional radar systems estimate target velocity primarily based on the Doppler effect, which yields only a one-dimensional radial velocity estimate. This limitation restricts overall radar performance, as only partial information about target motion can be estimated. To overcome this limitation, estimating the lateral velocity in addition to the radial velocity enables the recovery of the two-dimensional velocity vector, thereby significantly enhancing radar sensing capability. To this end, a beat signal model for a single-input multiple-output frequency-modulated continuous wave radar is derived using a second-order approximation, and the phase terms associated with lateral velocity are identified. Based on this model, a lateral velocity estimation method is proposed, in which individual point target signals are extracted and the discrete polynomial-phase transform is applied to obtain a two-dimensional lateral velocity spectrum. Furthermore, an enhanced radial velocity estimation method is developed to mitigate undesirable phase terms induced by lateral motion. Simulation results demonstrate that the proposed velocity vector estimation method can successfully estimate the velocity vectors of multiple targets. Statistical performance analysis is presented to validate the effectiveness of the proposed method under various conditions.