Research on Network Intrusion Detection Based on Support Vector Machine Optimized with Pigeon-inspired Optimization Algorithm
Yiheng Sun, Zhiwei Ye, Chunzhi Wang, Lingyu Yan, Ruoxi Wang · 2018
As an important technology in the field of network and information security, intrusion detection plays an important role in the information security protection system, and support vector machine (SVM) is one of the most successful methods. However, the performance of SVM is affected by its parameters. In order to improve the network intrusion detection accuracy, pigeon-inspired optimization (PIO) is introduced into intrusion detection to optimize the SVM parameters (PIO-SVM). PIO-SVM uses network intrusion detection data as the inputs of SVM, and SVM parameters are optimized by pigeon individual in PIO, the network intrusion detection accuracy is used as PIO target function, and then through mutual cooperation between pigeons, SVM parameters are obtained. Finally, the optimal model is used to detect the network intrusion detection. PIO-SVM is tested with KDDcup99 network intrusion data by using Matlab. The experimental results show that the detection accuracy of PIO-SVM is better than GA-SVM and PSO-SVM, which is a practical method for network intrusion detection.