Yanghui Xu
Yanghui Xu has a total of 10 co-authors. Here are the top co-authors by number of shared publications:
- Junxiong Guo
School of Electronic Information and Electrical Engineering, Chengdu University, Chengdu 610106, China
3 shared publications - Yingjun Chu
The Fifth Research Institute of Telecommunications Science and Technology Co., Ltd., Chengdu 610021, China
3 shared publications - Zhenghua Luo
School of Electronic Information and Electrical Engineering, Chengdu University, Chengdu 610106, China
3 shared publications - Shuyi Gu
Southeast University
2 shared publications - Chao Zeng
School of Electronic Information and Electrical Engineering, Chengdu University, Chengdu 610106, China
1 shared publication - Fuhao Lu
School of Electronic Information and Electrical Engineering, Chengdu University, Chengdu 610106, China
1 shared publication
Latest Publications
- Open Access
Real-Time Detection Sensor for Unmanned Aerial Vehicle Using an Improved YOLOv8s Algorithm
This study advances the unmanned aerial vehicle (UAV) localization technology within the framework of a low-altitude economy, with particular emphasis on the accurate and real-time identification and tracking of unauthorized (“black-flying”) drones. Conventional YOLOv8s-based target detection algorithms often suffer from missed detections due to their reliance on single-frame features. To address this limitation, this paper proposes an improved detection algorithm that integrates a long-short-term memory (LSTM) network into the YOLOv8s framework. By incorporating time-series modeling, the LSTM module enables the retention of historical features and dynamic prediction of UAV trajectories. The loss function combines bounding box regression loss with binary cross-entropy and is optimized using the Adam algorithm to enhance training convergence. The training data distribution is validated through Monte Carlo random sampling, which improves the model’s generalization to complex scenes. Simulation results demonstrate that the proposed method significantly enhances UAV detection performance. In addition, when deployed on the RK3588-based embedded system, the method achieves a low false negative rate and exhibits robust detection capabilities, indicating strong potential for practical applications in airspace management and counter-UAV operations.
Open AccessReal-Time Detection Sensor for Unmanned Aerial Vehicle Using an Improved YOLOv8s Algorithm
This study advances the unmanned aerial vehicle (UAV) localization technology within the framework of a low-altitude economy, with particular emphasis on the accurate and real-time identification and tracking of unauthorized (“black-flying”) drones. Conventional YOLOv8s-based target detection algorithms often suffer from missed detections due to their reliance on single-frame features. To address this limitation, this paper proposes an improved detection algorithm that integrates a long-short-term memory (LSTM) network into the YOLOv8s framework. By incorporating time-series modeling, the LSTM module enables the retention of historical features and dynamic prediction of UAV trajectories. The loss function combines bounding box regression loss with binary cross-entropy and is optimized using the Adam algorithm to enhance training convergence. The training data distribution is validated through Monte Carlo random sampling, which improves the model’s generalization to complex scenes. Simulation results demonstrate that the proposed method significantly enhances UAV detection performance. In addition, when deployed on the RK3588-based embedded system, the method achieves a low false negative rate and exhibits robust detection capabilities, indicating strong potential for practical applications in airspace management and counter-UAV operations.
- Open Access
- 1 September 2023
- Published by MDPI AG in Electronics
- Vol. 12 (17) , 3696
Research and Implementation of High-Efficiency and Low-Complexity LDPC Coding Algorithm
In this work, we proposed a high-efficiency and low-complexity encoding algorithm and its corresponding implementation structure during the design and implementation process of an LDPC encoder and decoder. This proposal was derived from extensive research on and analysis of standard encoding algorithms and recursive iterative encoding algorithms, specifically targeting the problem of high computational complexity in encoding algorithms. Subsequently, we combined binary phase-shift keying modulation mode and additive white Gaussian noise channel transmission with the min-sum decoding algorithm to realize the (1536, 1024) LDPC codec. This codec was uniformly quantized with a (6, 2) configuration, executed eight iterations, and achieved a 2/3 code rate in the IEEE802.16e standard. At the bit error rate (BER) of , the codec’s BER obtained by the proposed coding algorithm was about 0.25 dB lower than the recursive-iterative coding algorithm and was about 1.25 dB lower than the standard coding algorithm, which confirms the correctness, effectiveness, and feasibility of the proposed algorithm.- 1 September 2023
- Published by MDPI AG in Electronics
- Vol. 12 (17) , 3696
Open AccessResearch and Implementation of High-Efficiency and Low-Complexity LDPC Coding Algorithm
In this work, we proposed a high-efficiency and low-complexity encoding algorithm and its corresponding implementation structure during the design and implementation process of an LDPC encoder and decoder. This proposal was derived from extensive research on and analysis of standard encoding algorithms and recursive iterative encoding algorithms, specifically targeting the problem of high computational complexity in encoding algorithms. Subsequently, we combined binary phase-shift keying modulation mode and additive white Gaussian noise channel transmission with the min-sum decoding algorithm to realize the (1536, 1024) LDPC codec. This codec was uniformly quantized with a (6, 2) configuration, executed eight iterations, and achieved a 2/3 code rate in the IEEE802.16e standard. At the bit error rate (BER) of , the codec’s BER obtained by the proposed coding algorithm was about 0.25 dB lower than the recursive-iterative coding algorithm and was about 1.25 dB lower than the standard coding algorithm, which confirms the correctness, effectiveness, and feasibility of the proposed algorithm. - 30 August 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Information Forensics and Security
- Vol. 18 (15566013) , 5580-5591
Trace Alignment Preprocessing in Side-Channel Analysis Using the Adaptive Filter
Trace alignment can improve the subsequent side-channel analysis against the trace. Most current trace alignment schemes are, however, typically operated under a high signal-to-noise ratio (SNR), which demands them to be noise reduced before alignment when practical applications in the complex environment. In this paper, we propose a novel strategy for applying adaptive filtering in trace alignment preprocessing under low SNR conditions. The approach selects a trace as the reference signal of the adaptive filter, and the impulse response describing the trace offset is calculated iteratively for each trace. Different from conventional trace alignment methods, the error between the two traces in iteration determines how to eliminate the offset between trace, which eliminate most of the noise effects in the iteration process. In parallel, the filter after iterating will also function as a low-pass filter in the alignment process. Experimental studies based on three side-channel datasets demonstrate the efficacy of the proposed approach. Compared with other alignment methods, with the reasonable computational resource cost and complexity, the average number of traces required has reduced the average number of traces required by 75%, the average confidence has improved by 60%, and the success rate has increased by 72%. Our approach provides great potential for applications in trace alignment preprocessing of side-channel analysis.
- 30 August 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Information Forensics and Security
- Vol. 18 (15566013) , 5580-5591
Trace Alignment Preprocessing in Side-Channel Analysis Using the Adaptive Filter
Trace alignment can improve the subsequent side-channel analysis against the trace. Most current trace alignment schemes are, however, typically operated under a high signal-to-noise ratio (SNR), which demands them to be noise reduced before alignment when practical applications in the complex environment. In this paper, we propose a novel strategy for applying adaptive filtering in trace alignment preprocessing under low SNR conditions. The approach selects a trace as the reference signal of the adaptive filter, and the impulse response describing the trace offset is calculated iteratively for each trace. Different from conventional trace alignment methods, the error between the two traces in iteration determines how to eliminate the offset between trace, which eliminate most of the noise effects in the iteration process. In parallel, the filter after iterating will also function as a low-pass filter in the alignment process. Experimental studies based on three side-channel datasets demonstrate the efficacy of the proposed approach. Compared with other alignment methods, with the reasonable computational resource cost and complexity, the average number of traces required has reduced the average number of traces required by 75%, the average confidence has improved by 60%, and the success rate has increased by 72%. Our approach provides great potential for applications in trace alignment preprocessing of side-channel analysis.
- Open Access
A Suboptimal Optimizing Strategy for Velocity Vector Estimation in Single-Observer Passive Localization
In a single-observer passive localization system, the velocity and position of the target are estimated simultaneously. However, this can lead to correlated errors and distortion of the estimated value, making independent estimation of the speed and position necessary. In this study, we introduce a novel optimization strategy, suboptimal estimation, for independently estimating the velocity vector in single-observer passive localization. The suboptimal estimation strategy converts the estimation of the velocity vector into a search for the global optimal solution by dynamically weighting multiple optimization criteria from the starting point in the solution space. Simulation verification is conducted using uniform motion and constant acceleration models. The results demonstrate that the proposed method converges faster with higher accuracy and strong robustness.
Open AccessA Suboptimal Optimizing Strategy for Velocity Vector Estimation in Single-Observer Passive Localization
In a single-observer passive localization system, the velocity and position of the target are estimated simultaneously. However, this can lead to correlated errors and distortion of the estimated value, making independent estimation of the speed and position necessary. In this study, we introduce a novel optimization strategy, suboptimal estimation, for independently estimating the velocity vector in single-observer passive localization. The suboptimal estimation strategy converts the estimation of the velocity vector into a search for the global optimal solution by dynamically weighting multiple optimization criteria from the starting point in the solution space. Simulation verification is conducted using uniform motion and constant acceleration models. The results demonstrate that the proposed method converges faster with higher accuracy and strong robustness.