Xiaobo Yang
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Research Interests
Xiaobo Yang has a total of 46 co-authors. Here are the top co-authors by number of shared publications:
- Guolong Cui
University of Electronic Science and Technology of China
15 shared publications - Shisheng Guo
University of Electronic Science and Technology of China
13 shared publications - Lingjiang Kong
University of Electronic Science and Technology of China
8 shared publications - Jiahui Chen
University of Electronic Science and Technology of China
5 shared publications - Peilun Wu
School of Information and Communication Engineering, University of Electronic Science and Technology of China, Chengdu, China
4 shared publications - Zhihao Zhu
School of Information and Communication Engineering, University of Electronic Science and Technology of China, Chengdu, China
4 shared publications
Latest Publications
- 1 November 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Intelligent Transportation Systems
- Vol. 27 (11) , 15147-15163
Spatio-Temporal Dual-Branch Fusion Network for Multi-Frame Automotive Radar Target Detection
- 1 November 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Intelligent Transportation Systems
- Vol. 27 (11) , 15147-15163
Spatio-Temporal Dual-Branch Fusion Network for Multi-Frame Automotive Radar Target Detection
- 1 August 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Sensors Journal
- Vol. 26 (16) , 24523-24535
Radar-Based Continuous Human Activity Recognition With Complex-Valued Time--Frequency Data Fusion
- 1 August 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Sensors Journal
- Vol. 26 (16) , 24523-24535
Radar-Based Continuous Human Activity Recognition With Complex-Valued Time--Frequency Data Fusion
- 1 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Aerospace and Electronic Systems
- Vol. 62, 17385-17403
Waveform-Encrypted LPI Radar Framework: Joint Sequence Design, Robust Decryption, and Analysis
- 1 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Aerospace and Electronic Systems
- Vol. 62, 17385-17403
Waveform-Encrypted LPI Radar Framework: Joint Sequence Design, Robust Decryption, and Analysis
- 1 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Aerospace and Electronic Systems
- Vol. 62, 14918-14933
FIT: Feasibility-Informed mmWave Target Tracking via Multisource Spatial Constraints
- 1 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Aerospace and Electronic Systems
- Vol. 62, 14918-14933
FIT: Feasibility-Informed mmWave Target Tracking via Multisource Spatial Constraints
- 23 December 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Aerospace and Electronic Systems
- Vol. 62 (00189251) , 3569-3587
Multipath Exploitation-Based 3-D Environmental Perception and NLOS Moving Target Reconstruction for mmWave MIMO Imaging Radar
Multipath exploitation techniques have attracted considerable attention over the past decade, as they expand the detection coverage of conventional radar systems to Non-Line-of-Sight (NLOS) areas. However, existing methods rely on fully or partially known reflector surface information to achieve detection and localization of NLOS targets, which poses challenges for practical applications since scene information is often not available a priori. To overcome this issue, we present a multipath exploitation-based millimeter-wave (mmWave) multiple-input multiple-output (MIMO) imaging radar framework, which first performs three-dimensional (3D) environmental perception and then transfers the reflector surface parameters to assist NLOS target reconstruction. To implement this process, we explore several key problems and propose corresponding solutions. First of all, we propose a joint multiple linear array beamforming (JMLA-BF)-based imaging algorithm to achieve high-resolution sparse imaging as well as extract weak echoes of NLOS reflector surface. Then, an iteration algorithm is proposed for NLOS building layout recognition and relocation. Meanwhile, we propose an unambiguous direction-of-departure (DoD) and direction-of-arrival (DoA) estimation method for multipath signals from NLOS moving target using colocated MIMO radar. Finally, a path-oriented strategy is introduced to achieve NLOS target localization. Real experimental results demonstrate the effectiveness of the proposed framework and solutions in accurate 3D environmental perception and reliable NLOS target reconstruction.
- 23 December 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Aerospace and Electronic Systems
- Vol. 62 (00189251) , 3569-3587
Multipath Exploitation-Based 3-D Environmental Perception and NLOS Moving Target Reconstruction for mmWave MIMO Imaging Radar
Multipath exploitation techniques have attracted considerable attention over the past decade, as they expand the detection coverage of conventional radar systems to Non-Line-of-Sight (NLOS) areas. However, existing methods rely on fully or partially known reflector surface information to achieve detection and localization of NLOS targets, which poses challenges for practical applications since scene information is often not available a priori. To overcome this issue, we present a multipath exploitation-based millimeter-wave (mmWave) multiple-input multiple-output (MIMO) imaging radar framework, which first performs three-dimensional (3D) environmental perception and then transfers the reflector surface parameters to assist NLOS target reconstruction. To implement this process, we explore several key problems and propose corresponding solutions. First of all, we propose a joint multiple linear array beamforming (JMLA-BF)-based imaging algorithm to achieve high-resolution sparse imaging as well as extract weak echoes of NLOS reflector surface. Then, an iteration algorithm is proposed for NLOS building layout recognition and relocation. Meanwhile, we propose an unambiguous direction-of-departure (DoD) and direction-of-arrival (DoA) estimation method for multipath signals from NLOS moving target using colocated MIMO radar. Finally, a path-oriented strategy is introduced to achieve NLOS target localization. Real experimental results demonstrate the effectiveness of the proposed framework and solutions in accurate 3D environmental perception and reliable NLOS target reconstruction.
- 9 December 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Aerospace and Electronic Systems
- Vol. 62 (00189251) , 2883-2899
MDTrack: Micro-Doppler-Assisted Multitarget Tracking via Siamese Network Using MmWave Radar
Target tracking in dynamic environments is a critical component of radar-based indoor system, particularly for multi-target scenarios where maintaining identity consistency is a challenging task. In this paper, we propose the micro-doppler assisted multi-target tracking method (MDTrack), a millimeter-wave (mmWave) radar-based tracking algorithm that integrates micro-Doppler($\mu$D) features for effective track re-association and identity preservation. The method begins by extracting $\mu$D spectrograms of coexisting targets from radar data and applying a dual-threshold-based filtering method to assess the quality of the spectrograms. A siamese deep convolutional neural network, named S-DCNN, is then used to evaluate identity consistency between track segments, enabling re-identification without relying on explicit identity labels. Next, MDTrack employs a trajectory re-association approach that utilizes the temporal consistency of $\mu$D features, ensuring stable target tracking even in the presence of occlusions and trajectory fragmentation. Experimental results demonstrate that MDTrack significantly outperforms traditional methods, achieving a 94.42% accuracy for identity consistency and a 97.35% multi-object tracking accuracy in multi-target scenarios, which indicate robust performance in both closed-set and open-set conditions.
- 9 December 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Aerospace and Electronic Systems
- Vol. 62 (00189251) , 2883-2899
MDTrack: Micro-Doppler-Assisted Multitarget Tracking via Siamese Network Using MmWave Radar
Target tracking in dynamic environments is a critical component of radar-based indoor system, particularly for multi-target scenarios where maintaining identity consistency is a challenging task. In this paper, we propose the micro-doppler assisted multi-target tracking method (MDTrack), a millimeter-wave (mmWave) radar-based tracking algorithm that integrates micro-Doppler($\mu$D) features for effective track re-association and identity preservation. The method begins by extracting $\mu$D spectrograms of coexisting targets from radar data and applying a dual-threshold-based filtering method to assess the quality of the spectrograms. A siamese deep convolutional neural network, named S-DCNN, is then used to evaluate identity consistency between track segments, enabling re-identification without relying on explicit identity labels. Next, MDTrack employs a trajectory re-association approach that utilizes the temporal consistency of $\mu$D features, ensuring stable target tracking even in the presence of occlusions and trajectory fragmentation. Experimental results demonstrate that MDTrack significantly outperforms traditional methods, achieving a 94.42% accuracy for identity consistency and a 97.35% multi-object tracking accuracy in multi-target scenarios, which indicate robust performance in both closed-set and open-set conditions.
- 24 September 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Internet of Things Journal
- Vol. 12 (23) , 51467-51482
Human Activity Recognition Based on Multipath Fusion in Non-Line-of-Sight Corner
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-LOS through-wall (NLOS-TW) scenarios, neglecting consideration for non-LOS corner (NLOS-C) scenario within urban architecture. In the NLOS-C scenario, electromagnetic waves illuminate the target through multiple paths, resulting in considerable variations in the range–time map, causing performance degradation or even failure of existing methods. Moreover, multipath propagation enables a single-node radar to function as a multiperspective multinode radar system, providing more comprehensive information for human activity. Therefore, considering the complementary interpretations of multipath and the distinctive features observed in the NLOS-C range–time map, this article proposes a HAR method for the NLOS-C scenario based on multipath fusion. First, considering the broad distribution and large span characteristics of behavior features in the range dimension caused by the multipath effect, we design a multipath information fusion module based on dilated convolution to effectively integrate and interact with the multipath information. In addition, to address the diverse feature scales caused by variable widths and blurred boundaries of each path, we incorporate a multiscale unit into the deep feature extraction module to enhance the capability of autonomously adjusting the receptive field. Finally, multipath interaction information is fused with depth features for recognition. Experimental results validate the effectiveness of the proposed method in the NLOS-C scenario. The codes are available at: https://github.com/tlz1111/Multipath-Fusion-Network
- 24 September 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Internet of Things Journal
- Vol. 12 (23) , 51467-51482
Human Activity Recognition Based on Multipath Fusion in Non-Line-of-Sight Corner
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-LOS through-wall (NLOS-TW) scenarios, neglecting consideration for non-LOS corner (NLOS-C) scenario within urban architecture. In the NLOS-C scenario, electromagnetic waves illuminate the target through multiple paths, resulting in considerable variations in the range–time map, causing performance degradation or even failure of existing methods. Moreover, multipath propagation enables a single-node radar to function as a multiperspective multinode radar system, providing more comprehensive information for human activity. Therefore, considering the complementary interpretations of multipath and the distinctive features observed in the NLOS-C range–time map, this article proposes a HAR method for the NLOS-C scenario based on multipath fusion. First, considering the broad distribution and large span characteristics of behavior features in the range dimension caused by the multipath effect, we design a multipath information fusion module based on dilated convolution to effectively integrate and interact with the multipath information. In addition, to address the diverse feature scales caused by variable widths and blurred boundaries of each path, we incorporate a multiscale unit into the deep feature extraction module to enhance the capability of autonomously adjusting the receptive field. Finally, multipath interaction information is fused with depth features for recognition. Experimental results validate the effectiveness of the proposed method in the NLOS-C scenario. The codes are available at: https://github.com/tlz1111/Multipath-Fusion-Network
- 28 August 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Aerospace and Electronic Systems
- Vol. 61 (6) , 16474-16491
Interrupted-Sampling Repeater Jamming Recognition Based on PC-LFM Waveform Processing
The interrupted-sampling repeater jamming (lSRJ) generates a series of false targets after pulse compression by quickly sampling and forwarding radar signals, which poses a significant threat to radar performance. In order to facilitate the elimination of the periodic forwarding compound ISRJs, this paper focuses on discriminating between targets and ISRJ based on the phase-coded linear frequency modulation (PC-LFM) waveform. At the transmission stage, the peak sidelobe level (PSL) of the PC-LFM waveform's auto-correlation is minimized using an integrated Riemannian manifold optimization solver to enhance target detection capabilities. Taking advantage of the energy-reduction effect of the decoding operation in the PC-LFM on non-detection delay signals, a feature extraction scheme is developed to explore the disparity characteristic between the target and the ISRJ in the range profile. Subsequently, the PC-LFM-based recognition (PCLFM-R) algorithm is provided based on the feature difference in the number of peaks and the width of the mainlobe. Moreover, decoding filters are designed to improve recognition accuracy further, leading to a modified PCLFM-R (M-PCLFM-R) procedure. Numerical results demonstrate the effectiveness of the proposed PCLFM-R and M-PCLFM-R approaches in discrimination targets and ISRJ under several aliasing interferences.
- 28 August 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Aerospace and Electronic Systems
- Vol. 61 (6) , 16474-16491
Interrupted-Sampling Repeater Jamming Recognition Based on PC-LFM Waveform Processing
The interrupted-sampling repeater jamming (lSRJ) generates a series of false targets after pulse compression by quickly sampling and forwarding radar signals, which poses a significant threat to radar performance. In order to facilitate the elimination of the periodic forwarding compound ISRJs, this paper focuses on discriminating between targets and ISRJ based on the phase-coded linear frequency modulation (PC-LFM) waveform. At the transmission stage, the peak sidelobe level (PSL) of the PC-LFM waveform's auto-correlation is minimized using an integrated Riemannian manifold optimization solver to enhance target detection capabilities. Taking advantage of the energy-reduction effect of the decoding operation in the PC-LFM on non-detection delay signals, a feature extraction scheme is developed to explore the disparity characteristic between the target and the ISRJ in the range profile. Subsequently, the PC-LFM-based recognition (PCLFM-R) algorithm is provided based on the feature difference in the number of peaks and the width of the mainlobe. Moreover, decoding filters are designed to improve recognition accuracy further, leading to a modified PCLFM-R (M-PCLFM-R) procedure. Numerical results demonstrate the effectiveness of the proposed PCLFM-R and M-PCLFM-R approaches in discrimination targets and ISRJ under several aliasing interferences.
- 13 June 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Aerospace and Electronic Systems
- Vol. 61 (5) , 13647-13661
A Non-Line-of-Sight Human Activity Recognition Method Based on Radar Multispectrogram
In this paper, we investigate a lightweight human activity recognition (HAR) method for non-line-of-sight (NLOS) scenarios. This approach constructs a lightweight multi-spectrogram parallel feature extraction model to extract time-range (TR) map and time-Doppler (TD) map features from stepped-frequency continuous wave (SFCW) radar, enabling NLOS HAR in L-shaped corner environments. Specifically, the model incorporates a multiscale depthwise separable convolution to extract multipath TR map features at different scales in NLOS scenarios. Simultaneously, it combines depthwise separable convolution with Ghost convolution to extract features from the TD map. Finally, the extracted multispectrogram features are fused and classified, achieving robust HAR in NLOS environments. To validate the effectiveness of the proposed method, we conducted experimental verification using real-world data from eight types of human activities in an L-shaped corner NLOS scenario. The results show that the proposed model achieves a recognition accuracy of 87.12%, which is a 5.82% improvement compared to the state-of-the-art recognition methods.
- 13 June 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Aerospace and Electronic Systems
- Vol. 61 (5) , 13647-13661
A Non-Line-of-Sight Human Activity Recognition Method Based on Radar Multispectrogram
In this paper, we investigate a lightweight human activity recognition (HAR) method for non-line-of-sight (NLOS) scenarios. This approach constructs a lightweight multi-spectrogram parallel feature extraction model to extract time-range (TR) map and time-Doppler (TD) map features from stepped-frequency continuous wave (SFCW) radar, enabling NLOS HAR in L-shaped corner environments. Specifically, the model incorporates a multiscale depthwise separable convolution to extract multipath TR map features at different scales in NLOS scenarios. Simultaneously, it combines depthwise separable convolution with Ghost convolution to extract features from the TD map. Finally, the extracted multispectrogram features are fused and classified, achieving robust HAR in NLOS environments. To validate the effectiveness of the proposed method, we conducted experimental verification using real-world data from eight types of human activities in an L-shaped corner NLOS scenario. The results show that the proposed model achieves a recognition accuracy of 87.12%, which is a 5.82% improvement compared to the state-of-the-art recognition methods.