Open AccessLSH-XGBoost based Network Congestion Detection Method for SSDN
Author(s) -
Kaiyuan Tian,
Jian Wang,
Yuanyuan Liao,
Dengke Xu,
Baigen Cai
Publication year - 2020
Publication title -
journal of physics. conference series
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.21
H-Index - 85
eISSN - 1742-6596
pISSN - 1742-6588
DOI - 10.1088/1742-6596/1549/5/052069
Subject(s) - computer science , denial of service attack , intrusion detection system , real time computing , boosting (machine learning) , transmission (telecommunications) , traffic congestion , computer network , data mining , artificial intelligence , telecommunications , engineering , operating system , transport engineering , the internet
Signal Safety Data Network (SSDN) is essential to maintain the reliable transmission between station and station equipment, station and central signal equipment. Threats in SSDN include but not limited to poor performance in tele-communication and denial of service when network congestion occurs. Therefore, it’s of great significance detecting the potential network congestion in railroad SSDN. Existing detection algorithms are low efficient, less accurate and are unable to cover all the detection categories. By combining the advantages of low time complexity and high classification accuracy originate from Local Sensitive Hashing (LSH) and eXtreme Gradient Boosting (XGBoost), we introduce a network congestion detection method based on LSH-XGBoost for SSDN. We run testing on open source intrusion detection system: Snort, which was built on SSDN simulator. The promising result proved its capability.