A5053674695- Network Security and Intrusion Detection
- Internet Traffic Analysis and Secure E-voting
- Network Packet Processing and Optimization
- Advanced Malware Detection Techniques
- Algorithms and Data Compression
- Nonlinear Dynamics and Pattern Formation
- Vibration and Dynamic Analysis
- Chaos control and synchronization
- Anomaly Detection Techniques and Applications
- Quantum Computing Algorithms and Architecture
- Advanced Authentication Protocols Security
- Caching and Content Delivery
- Quantum Information and Cryptography
- Quantum chaos and dynamical systems
- Fluid Dynamics and Vibration Analysis
- Spam and Phishing Detection
- Cryptography and Data Security
- Text and Document Classification Technologies
- Quantum-Dot Cellular Automata
- Hydrogen embrittlement and corrosion behaviors in metals
- Material Properties and Failure Mechanisms
- Railway Engineering and Dynamics
- Metal and Thin Film Mechanics
- Hate Speech and Cyberbullying Detection
- Remote Sensing and Land Use
National University of Defense Technology
2016-2025
Keelung Chang Gung Memorial Hospital
2025
Wuhan University
2011-2024
Chongqing University
2024
University of Science and Technology Beijing
2024
China State Shipbuilding (China)
2024
Jiyang College of Zhejiang A&F University
2023
Central South University
2019-2022
Wuhan Textile University
2021-2022
Wuhan University of Technology
2022
Deep packet inspection (DPI) is widely used in content-aware network applications such as intrusion detection systems, traffic billing, load balancing, and government surveillance. Pattern matching a core critical step DPI, which checks the payload of each for known signatures (patterns) order to identify packets with certain characteristics (e.g., malicious that carry viruses or worms). Regular expression major tool signature description due its powerful flexible expressive ability....
The proliferation of mobile devices over recent years has led to a dramatic increase in traffic. Demand for enabling accurate app identification is coming as it an essential step improve multitude network services: accounting, security monitoring, traffic forecasting, and quality-of-service. However, traditional classification techniques do not work well Besides, multiple machine learning solutions developed this field are severely restricted by their handcrafted features unreliable...
Abstract An efficient cryptography scheme is proposed based on continuous-variable quantum neural network (CV-QNN), in which a specified CV-QNN model introduced for designing the algorithm. It indicates an approach to design cryptosystem contains processes of key generation, encryption and decryption. Security analysis demonstrates that our security. Several simulation experiments are performed Strawberry Fields platform processing classical data “Quantum Cryptography” with describe...
Purpose – This purpose of this study is to apply social exchange theory (SET) explain how behaviors, such as the knowledge, information and respect between firms, would increase likelihood certification implementation strengthen relationships among partners. The main purposes are examine significant connections partner interactions long-term orientation links orientation, certifications relationship quality. Design/methodology/approach To test research hypotheses, structural equation...
With the exponential growth of mobile traffic data, classification is in a great need. It an essential step to improve performance network services such as QoS and security monitoring. However, widespread use encrypted protocols, especially TLS protocol, has posed challenges traditional techniques. As rule-based deep packet inspection approaches are ineffective for classification, various machine learning methods have been studied used. Recently, solutions which enable automatic feature...
The botnet relies on the Command and Control (C&C) channels to conduct its malicious activities remotely. Domain Generation Algorithm (DGA) is often used by botnets hide their server evade take-down attempts, which allows bot generate a large number of domain names until it finds C&C server. lengths generated DGAs are different. Our research that length name has an impact performance DGA detection model. In other words, model sensitive name. this case, attackers can simply designing specific...
Objectives To assess the feasibility and acceptability of a randomized controlled trial compared auricular acupressure (AA) on specific acupoints with AA non-specific for treating maintenance hemodialysis (MHD) patients insomnia. Methods Sixty three (63) eligible subjects were randomly assigned into either group received (n=32), or sham (SAA) points irrelevant to insomnia treatment (n=31) eight weeks. All participants followed up 12 weeks after treatments. The primary outcome was clinical...
Abstract Graph neural network (GNN)-based fault diagnosis has attracted widespread attention. However, the existing measure metrics of domain distribution discrepancy always is single, weight different domains unknown, and scale GNN-based feature extractor unknown. To address these issues, a time graph sub-domain adaption adversarial (TGSDAA) for proposed in this paper. Firstly, multi-scale connection layer to extract signal. Specifically, an encode construct undirected graph. Next,...
The proliferation of handheld devices has led to an explosive growth mobile traffic volumes on the Internet. Identifying apps from network become a crucial task for management and security. Traditionally, design accurate identifiers relies deep packet inspection (DPI) techniques. However, such approaches have less effective with raising adoption encrypted protocols in applications (mostly TLS). To address problem, various machine learning methods been studied used. Most them use linear...
Semantic change detection, which aims to locate land cover changes and identify their categories using pixel-level boundaries, has promising applications in Earth vision, including precise urban planning natural resource management. This paper proposes a novel Siamese U-Net architecture for semantic detection synthetic aperture radar (SAR) images, incorporating residual network with weight-sharing as the backbone network. The is capable of simultaneously yielding binary results....
Abstract The microstructure of materials has a great influence on the macroscopic mechanical properties materials, and relationship between them is significance to design materials. With development artificial intelligence deep learning, many researchers have used convolutional neural networks (CNN) correlate microscopic images with material properties, achieved good results. However, for different types performance prediction tasks datasets sizes, need special CNN network structure,...
A sufficient criterion for synchronizing two identical non-autonomous chaotic systems coupled by sinusoidal state error feedback control is firstly derived in the form of a matrix inequality Lyapunov direct method. The then simplified into some algebraic inequalities based on Gerschgorin disc theorem. In order to obtain simpler criteria, variables are restricted sub-region. property that similar matrices have same eigenvalues further applied more flexible synchronization criteria. Numerical...
Regular expression matching engine is a crucial infrastructure which widely used in network security systems, like IDS. We propose Gregex, Graphics Processing Unit (GPU) based regular for deep packet inspection (DPI). Gregex leverages the computational power and high memory bandwidth of GPUs by storing data proper GPU space executing massive thread concurrently to process lots packets parallel. Three optimization techniques, ATP, CAB, CAT are proposed significantly improve performance...