A5079428812- Software-Defined Networks and 5G
- Internet Traffic Analysis and Secure E-voting
- Network Security and Intrusion Detection
- Vehicular Ad Hoc Networks (VANETs)
- Advanced Malware Detection Techniques
- Traffic Prediction and Management Techniques
- Network Traffic and Congestion Control
- Brain Tumor Detection and Classification
- IoT and Edge/Fog Computing
Wuhan University
2019-2022
While software defined network (SDN) brings more innovation to the development of future networks, it also faces a severe threat from DDoS attacks. In order deal with single point failure on SDN controller caused by attacks, we propose framework for detection and defense attacks in environment. Firstly, deploy trigger mechanism attack data plane screen abnormal flows network. Then, use combined machine learning algorithm based K-Means KNN exploit rate characteristics asymmetry detect...
With the development of data plane programmable Software-Defined Networking (SDN), Distributed Denial Service (DDoS) attacks on increasingly become fatal. Currently, traditional attack detection methods are mainly used to detect whether a DDoS occurs and it is difficult find path that flow traverses network, which makes accurately mitigate attacks. In this article, we propose method based Spatial-Temporal Graph Convolutional Network (ST-GCN) over SDN, maps network into graph. It senses state...
Vehicular ad hoc network (VANET) has become an accessible technology for improving road safety and driving experience, the problems of heterogeneity lack resources it faces have also attracted widespread attention. With development software-defined networking (SDN) multiaccess edge computing (MEC), a variety resource allocation strategies in MEC-enabled networking-based VANET (SDVN) been proposed to solve these problems. However, we note that few work involves situation where SDVN is under...
In the network, accurately predicting network flow queuing delay is important for congestion control, bandwidth allocation, and performance improvement. order to improve QoS (quality of service) predict fine-grained in advance, A framework proposed this paper. We firstly perform real-time preprocessing on data packet, which acquired using In-band Network Telemetry (INT) technology, based Storm platform. Then we use Spark Streaming online prediction LSTM short-term model obtained by training...
Deep learning algorithms have demonstrated remarkable efficacy in the medical imaging field, particularly when it comes to segmentation and classification of brain tumors. This algorithm is being trained tested using MRI tumor dataset provided by Kaggle. ResNet50 used for tasks due its powerful feature extraction capability. Through deep residual structure, can effectively extract different lesion features improve accuracy. At same time, ResUNet combines capabilities ResNet with advantages...