Xiaoru Song
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Xiaoru Song has a total of 4 co-authors. Here are the top co-authors by number of shared publications:
- Tao Zhang
School of Mechatronic Engineering, Xi’an Technological University, Xi’an 710021, China
3 shared publications - Chaobo Chen
School of Electronic Information Engineering, Xi’an Technological University, Xi’an 710021, China
1 shared publication - Song Gao
School of Electronic Information Engineering, Xi’an Technological University, Xi’an 710021, China
1 shared publication - Tianli Ma
School of Electronic Information Engineering, Xi’an Technological University, Xi’an 710021, China
1 shared publication
Latest Publications
- Open Access
Rotorcraft UAV Recognition Method Using Multi-Domain Feature Fusion Based on Millimeter-Wave Radar Sensor
To improve the recognition accuracy of rotorcraft unmanned aerial vehicles (UAVs) under millimeter-wave radar sensors, this paper proposes to use two millimeter-wave radars with the same wavelength band to detect UAVs. Based on the correlation of the UAV’s echo signals from two radars, we establish a fusion method of UAV echo signals detected by two radars based on the cross-correlation coefficient function method and develop a recognition algorithm for the echo signals of UAVs using multi-domain feature fusion measures. In the recognition algorithm, we process the fused echo signals of the UAV using the cross-correlation coefficient function; use time–frequency transformation to obtain time–frequency diagram, Cadence Velocity Diagram(CVD), and Cadence Frequency Spectrum (CFS) diagram; and then input the formed time–frequency diagram and CVD into the EfficientNet-B0-CBAM model and the CFS diagrams into the TCN-Transformer model and gain the multi-domain feature information of the UAV’s time–frequency, CVD, and CFS; lastly, we set up the multi-domain feature fusion extraction model to form a classification and recognition algorithm of the UAV’s echo signals. Based on our own built dataset of the UAV’s echoes, we conducted a comparative verification of the recognition algorithms of rotorcraft UAVs, and the results showed that the proposed method had an average improvement of 1.5% in terms of accuracy and precision.
Open AccessRotorcraft UAV Recognition Method Using Multi-Domain Feature Fusion Based on Millimeter-Wave Radar Sensor
To improve the recognition accuracy of rotorcraft unmanned aerial vehicles (UAVs) under millimeter-wave radar sensors, this paper proposes to use two millimeter-wave radars with the same wavelength band to detect UAVs. Based on the correlation of the UAV’s echo signals from two radars, we establish a fusion method of UAV echo signals detected by two radars based on the cross-correlation coefficient function method and develop a recognition algorithm for the echo signals of UAVs using multi-domain feature fusion measures. In the recognition algorithm, we process the fused echo signals of the UAV using the cross-correlation coefficient function; use time–frequency transformation to obtain time–frequency diagram, Cadence Velocity Diagram(CVD), and Cadence Frequency Spectrum (CFS) diagram; and then input the formed time–frequency diagram and CVD into the EfficientNet-B0-CBAM model and the CFS diagrams into the TCN-Transformer model and gain the multi-domain feature information of the UAV’s time–frequency, CVD, and CFS; lastly, we set up the multi-domain feature fusion extraction model to form a classification and recognition algorithm of the UAV’s echo signals. Based on our own built dataset of the UAV’s echoes, we conducted a comparative verification of the recognition algorithms of rotorcraft UAVs, and the results showed that the proposed method had an average improvement of 1.5% in terms of accuracy and precision.
- Open Access
A Classification Algorithm of UAV and Bird Target Based on L/K Dual-Band Micro-Doppler and Mamba
To address the challenge of accurately distinguishing UAVs and birds in a low-altitude detection field, this paper proposes a classification algorithm of UAVs and birds based on L/K dual-band micro-Doppler spectrograms and Mamba. We establish a dual-band radar detection model for unmanned aerial vehicles (UAVs) and birds, provide a method for characterizing the Doppler parameters of the echo signals, and research a UAV and bird target classification network model that integrates micro-Doppler and Mamba. Based on a dual-branch encoding framework, we use Patch block decomposition to design a classification model to serialize the two-dimensional spectrogram of the echo signal, and introduce the Mamba state-space backbone network to extract the long-term sequence features of the target’s micro-motion. The main breakthrough of the proposed classification algorithm lies in the feature fusion stage, where a late fusion strategy is adopted to integrate the dual-path high-level representation measures, fully leveraging the sensitivity of the K-band to high-frequency textures and the scale complementarity of the L-band. Then, according to the joint loss function of mutual learning and contrastive learning, we improve the model’s prediction consistency and representation discriminability. Through experimental testing, the results show that the proposed method can effectively classify UAVs and birds, and compared with other algorithms, the accuracy rate reaches 97.5%.
Open AccessA Classification Algorithm of UAV and Bird Target Based on L/K Dual-Band Micro-Doppler and Mamba
To address the challenge of accurately distinguishing UAVs and birds in a low-altitude detection field, this paper proposes a classification algorithm of UAVs and birds based on L/K dual-band micro-Doppler spectrograms and Mamba. We establish a dual-band radar detection model for unmanned aerial vehicles (UAVs) and birds, provide a method for characterizing the Doppler parameters of the echo signals, and research a UAV and bird target classification network model that integrates micro-Doppler and Mamba. Based on a dual-branch encoding framework, we use Patch block decomposition to design a classification model to serialize the two-dimensional spectrogram of the echo signal, and introduce the Mamba state-space backbone network to extract the long-term sequence features of the target’s micro-motion. The main breakthrough of the proposed classification algorithm lies in the feature fusion stage, where a late fusion strategy is adopted to integrate the dual-path high-level representation measures, fully leveraging the sensitivity of the K-band to high-frequency textures and the scale complementarity of the L-band. Then, according to the joint loss function of mutual learning and contrastive learning, we improve the model’s prediction consistency and representation discriminability. Through experimental testing, the results show that the proposed method can effectively classify UAVs and birds, and compared with other algorithms, the accuracy rate reaches 97.5%.
- Open Access
Computational Methods and Simulation of UAVs’ Micro-Motion Echo Characteristics Using Distributed Radar Detection
The large number of UAVs under supervision at low altitudes have brought serious security risks to the field of air defense. Accurately analyzing the characteristics of UAVs’ echo signals is of great research significance for the detection and recognition of UAVs. Based on the principle of radar detection, the echo spatial correlation in the distributed radar detection mode is studied. According to the influence of different movement speeds on the micro-motion characteristics of UAVs, the echo signal models of UAVs in two flight states are established. Combined with the instantaneous micro-Doppler frequency model of the ideal motion state of UAVs, micro-Doppler frequency calculation functions of UAVs at different attitude angles are constructed. Through simulation calculation, the variation curves between the observation angle and the echo spatial correlation using different detection distances are given. Based on time–frequency images of UAVs in their ideal motion state, changes in the time–frequency images at different motion speeds and attitude angles are analyzed. These research results will help radar detection systems to accurately recognize UAVs in an uncertain motion state and can also provide a basis for predicting the next motion action of UAVs in subsequent target tracking.
Open AccessComputational Methods and Simulation of UAVs’ Micro-Motion Echo Characteristics Using Distributed Radar Detection
The large number of UAVs under supervision at low altitudes have brought serious security risks to the field of air defense. Accurately analyzing the characteristics of UAVs’ echo signals is of great research significance for the detection and recognition of UAVs. Based on the principle of radar detection, the echo spatial correlation in the distributed radar detection mode is studied. According to the influence of different movement speeds on the micro-motion characteristics of UAVs, the echo signal models of UAVs in two flight states are established. Combined with the instantaneous micro-Doppler frequency model of the ideal motion state of UAVs, micro-Doppler frequency calculation functions of UAVs at different attitude angles are constructed. Through simulation calculation, the variation curves between the observation angle and the echo spatial correlation using different detection distances are given. Based on time–frequency images of UAVs in their ideal motion state, changes in the time–frequency images at different motion speeds and attitude angles are analyzed. These research results will help radar detection systems to accurately recognize UAVs in an uncertain motion state and can also provide a basis for predicting the next motion action of UAVs in subsequent target tracking.
- Open Access
Multitarget Tracking Algorithm Based on Adaptive Network Graph Segmentation in the Presence of Measurement Origin Uncertainty
To deal with the problem of multitarget tracking with measurement origin uncertainty, the paper presents a multitarget tracking algorithm based on Adaptive Network Graph Segmentation (ANGS). The multitarget tracking is firstly formulated as an Integer Programming problem for finding the maximum a posterior probability in a cost flow network. Then, a network structure is partitioned using an Adaptive Spectral Clustering algorithm based on the Nyström Method. In order to obtain the global optimal solution, the parallel A* search algorithm is used to process each sub-network. Moreover, the trajectory set is extracted by the Track Mosaic technique and Rauch–Tung–Striebel (RTS) smoother. Finally, the simulation results achieved for different clutter intensity indicate that the proposed algorithm has better tracking accuracy and robustness compared with the A* search algorithm, the successive shortest-path (SSP) algorithm and the shortest path faster (SPFA) algorithm.
Open AccessMultitarget Tracking Algorithm Based on Adaptive Network Graph Segmentation in the Presence of Measurement Origin Uncertainty
To deal with the problem of multitarget tracking with measurement origin uncertainty, the paper presents a multitarget tracking algorithm based on Adaptive Network Graph Segmentation (ANGS). The multitarget tracking is firstly formulated as an Integer Programming problem for finding the maximum a posterior probability in a cost flow network. Then, a network structure is partitioned using an Adaptive Spectral Clustering algorithm based on the Nyström Method. In order to obtain the global optimal solution, the parallel A* search algorithm is used to process each sub-network. Moreover, the trajectory set is extracted by the Track Mosaic technique and Rauch–Tung–Striebel (RTS) smoother. Finally, the simulation results achieved for different clutter intensity indicate that the proposed algorithm has better tracking accuracy and robustness compared with the A* search algorithm, the successive shortest-path (SSP) algorithm and the shortest path faster (SPFA) algorithm.