A5101632836- Network Security and Intrusion Detection
- Anomaly Detection Techniques and Applications
- Internet Traffic Analysis and Secure E-voting
- Network Traffic and Congestion Control
- Time Series Analysis and Forecasting
- Artificial Immune Systems Applications
- Image Processing Techniques and Applications
- Electromagnetic Compatibility and Noise Suppression
- Radio Frequency Integrated Circuit Design
- Microwave Engineering and Waveguides
- Fault Detection and Control Systems
- Cybersecurity and Information Systems
- Non-Invasive Vital Sign Monitoring
- Software-Defined Networks and 5G
- Advanced Data Processing Techniques
- Inertial Sensor and Navigation
Southeast University
2024
State Key Laboratory of Millimeter Waves
2024
Nanyang Technological University
2015-2018
Anomaly detection is critical given the raft of cyber attacks in wireless communications these days. It thus a challenging task to determine network anomaly more accurately. In this paper, we propose an Autoencoder-based method. Autoencoder able capture non-linear correlations between features so as increase accuracy. We also apply Convolutional (CAE) here perform dimensionality reduction. As has smaller number parameters, it requires less training time compared conventional Autoencoder. By...
In this paper, we propose a network anomaly detection system which consists of Maximal Information Coefficient based feature selection algorithm and feature-based MSPCA algorithm, can separate the anomalous information more efficiently. provide good measurement any dependency between two random variables. combines benefit PCA wavelet analysis to reduce effect normal subspace contamination, is main challenge PCA-based algorithm. We utilize multiple flow features describe traffic instead using...
DDoS attack detection has been given great attention in recent years. Some of these intrusion systems use statistical analysis method or wavelet transform at the core, but they both suffer from problem high false alarm rate. In this paper, we apply Multi-scale Principal Component Analysis (MSPCA) algorithm system which combines benefit PCA and analysis. Then based on behavior attack, generate several basic metrics to evaluate proposed MSPCA 1999 DARPA dataset. Our evaluation results show...
As most of consumer electronics are connected to the Internet, network attacks can cause massive damage and loss data users. By sending periodic packet bursts bottleneck routers, Low-Rate Denial-of-Service (LDoS) degrade throughput TCP applications while being hard be detected. In this paper, we introduce Power Spectrum Density Entropy (PSD-entropy) detect LDoS attacks. We also propose a Fourier transform based Robust RED (FRRED) queuing algorithm preserve when faced with This novel Active...
Anomaly detection is critical given the raft of cyber attacks these days. It thus essential to identify network anomalies more accurately. In this paper, we propose a novel anomaly system which combines random projections (sketches) and feature-based MSPCA detect anomalous source IP addresses. By combining PCA wavelet analysis, can separate data efficiently. Incorporating with Sketch structure enables our proposed system, extract several flow-based features are helpful in exposing different...
The paper presents an implementation of support for simulation SYN Flood DDoS attacks in NS-3 simulator. software architecture is outlined, as well code changes that were made. unbalanced dumbbell topology used simulation. simulated are presented visually using diagram Shannon entropy time series, calculated on distribution destination ports and addresses. enables researchers to study methods detection mitigation
In this paper, we address the oscillations in signals of both motion capture and inertial measurement sensors. This characteristic is often observed when range reaches or exceeds approximately 85 degrees. Elimination oscillation filtered output signal significance as it means can be applied directly by applications such visualization, tracking, clinical report, etc. paper proposed a system model using feature selection machine learning algorithm to automatically detect post-filter...