- Advanced Photonic Communication Systems
- Optical Wireless Communication Technologies
- Satellite Communication Systems
- Quantum Information and Cryptography
- Optical Network Technologies
- Radio Wave Propagation Studies
- Advanced Optical Network Technologies
- Optimization and Search Problems
- Industrial Vision Systems and Defect Detection
- Quantum Computing Algorithms and Architecture
- Radar Systems and Signal Processing
- Mobile Agent-Based Network Management
- Advanced Measurement and Detection Methods
- Remote Sensing and LiDAR Applications
- Quantum Mechanics and Applications
- Remote Sensing and Land Use
- Optical Systems and Laser Technology
- Reliability and Maintenance Optimization
- Face and Expression Recognition
- Advanced Fiber Laser Technologies
- Advanced Neural Network Applications
- Advanced Optical Sensing Technologies
- Photonic Crystal and Fiber Optics
- Semiconductor Lasers and Optical Devices
- Elevator Systems and Control
National University of Defense Technology
2018-2024
China Institute Of Communications
2015-2018
Air Force Engineering University
2012-2014
Nanjing University of Chinese Medicine
2013
The electrical connector is an essential component of all kinds electronic equipment. It important to conduct the remaining useful life (RUL) prediction connectors. intermittent fault phenomenon has been observed during degradation process This article proposes combining contact resistance features and faults predict RUL evolution characteristics in whole-life connectors are compared analyzed. found that precursor feature comes out earlier than rise resistance, which used determine first...
The faulty-free operation of the optical cable lines is key to stable military communication network. Because traditional manual hiking inspection and maintenance method for not efficient, time-consuming laborious, it proposed use Unmanned Aerial Vehicle (UAV) inspect lines. deep learning Faster R-CNN target detection algorithm applied identification hidden dangers in aerial image captured by UAV. danger refers excavation damage engineering vehicle. Based on Detection Imagery (VEDAI)...
Measurement-device-independent quantum key distribution is immune from all the detection attacks, thus when it combined with decoy state method, final unconditional secure. In this paper, performance of three-intensity measurement-device-independent at an asymmetric channel transmittance efficiency considered and compared each other symmetric choice scenario. Simulation result shows that rate scenario can tolerate 62 dB loss, otherwise distance ratio changes, tolerated loss will decrease to...
Aiming at the problem of oversubscription data relay access request user stars in future Space-Based Information System, resource scheduling optimization for satellite system with microwave and laser hybrid links is studied. The characteristics are analyzed. A multi-objective programming model on static constraint satisfaction established, a algorithm integration artificial immune strategies, niche ideas improved genetic put forward to solve model. Simulation results show that optimizes...
Measurement-device-independent quantum key distribution (MDI-QKD) is immune to all the detection attacks, thus when it combined with decoy state method, final unconditionally safe. In this paper, we propose perform MDI-QKD odd coherent (OCS) and compare results weak source scenario. Our simulation indicates that both secure rate transmission distance can be improved evidently OCS owing lower probability of multi-photon events OCS. Furthermore, apply finite analysis obtain a practical rate.
The original measurement of device-independent(MDI)quantum key distribution(QKD) is reviewed, and a modified protocol using rotation-invariant photonic state proposed. Initial encoding final decoding information in our MDI-QKD implementation can be conveniently performed the polarization space, while transmission done hybrid space. Our analysis indicates that both secure rate distance improved by owing to its lower error rate. Furthermore, polarization-OAM qubits approach only needs insert...
In this paper, the problem of tracking vehicles using lidar sensors mounted on an ego-vehicle is addressed. Due to line-of-sight limitations, back (or front) a vehicle as seen by behind ahead of) it often modeled L-shaped. L-shape based algorithm with joint kinematic and geometric estimation presented. By feeding results fitting, detection method that robust outliers proposed. tracker currently available in literature, states model are separately estimated maintained. However, not...
Measurement-device-independent quantum key distribution (MDI-QKD) is immune to all the detection attacks, thus when it combined with decoy-state method, final rate can be obtained by estimating gain and bit error for various input photon numbers. In this paper, we propose perform MDI-QKD odd coherent state (OCS) compare results weak source scenario. Our simulation indicates that both secure transmission distance improved evidently OCS owing lower probability of multi-photon events OCS....
Airborne optical communication systems have been widely used for high-speed data transmission. In this paper, the average bit error rate (BER) of M-ary pulse position modulation (PPM) airborne is analyzed under combined influence atmospheric turbulence and aero-optic effects. The modeled by exponentiated Weibull (EW) distribution that can accurately describe probability density function irradiance weak to strong conditions in presence aperture averaging. Next, we use our derived form study...
Since the construction of engineering vehicles such as excavators in vicinity optical cable lines is most important cause fault lines, it proposed to apply deep learning YOLOv3 (You Only Look Once version 3) target detection algorithm vehicle and warning aerial inspection image. Based on Darknet framework, rapid realized by making datasets. The speed 60ms per sheet, recognition AP (Average Precision) value 0.869, which exceeds algorithms Faster R-CNN (Faster Region-based Convolutional Neural...
Optical satellite communication with the advantages of broadband, large capacity and low power consuming broke bottleneck traditional microwave communication. The formation Space-based Information System technology high performance optical inter-satellite realization global seamless coverage mobile terminal accessing are necessary trend development Considering resources, missions restraints Data Relay Satellite Communication System, a model resources scheduling is established algorithm based...
Due to clutter inhomogeneity, the suppression ability of space–time adaptive processing (STAP) is usually constrained by insufficient number independent and identically distributed (IID) training samples and, as a result, sacrificed achieve demanded sample reduction. Moreover, since heterogeneity exacerbated in real environment, IID size can be heavily reduced, leading deterioration suppression. To solve this problem, novel robust joint sparse method with airborne active array proposed. This...
Abstract Due to the susceptibility of intelligent recognition systems adversarial sample attacks, research on attack methods for target detection is great significance improving safety vehicle autonomous driving. By studying transfer-based self-ensemble generation algorithm, and based trained Faster R-CNN YOLOX white box detectors, a cross-model cross-instance method was designed. Through training, universal patch obtained, which achieved attacks two types detectors three vehicles, resulting...
Intermittent faults are widely present in aviation electronic devices, especially various electrical connectors. It is usually hard to diagnose the source of intermittent faults, which brings a huge challenge repair and maintenance equipment. This paper focuses on typical connectors activated by shock test. The transient arc wave observed nanosecond scale during occurrence faults. An signal model constructed analyze impact factors signal. Based characteristics, further fault diagnostic...
Object detection and tracking are critical fundamental problems in machine vision task. In this paper, an object method is proposed based on deep feature distillation. Particularly, adaptive unsupervised Teacher-Student unified framework developed. The Teacher module performed by expandable generative adversarial network mixture model. And knowledge discrepancy ranking (KDR) designed to optimize resource allocation with the historical underlying knowledge. Student developed a lightweight...