Research On Wireless Intrusion Detection Technology Based On Openwrt

As an important part of wireless networks,intelligent terminal equipment often faces the threat of network attacks.However,the intelligent terminal device itself does not have security defense capabilities,and is easy to be attacked.Moreover,there are a large number of intelligent terminal devices,which makes it difficult to embed a security device in each intelligent terminal device.The router is the collection point of wireless network traffic,through which all intelligent terminal devices exchange information.Because of the limitations of traditional routers,it is difficult to develop a data capture system on them,but wireless routers equipped with the OpenWrt system have secondary development capabilities.Therefore,research on intrusion detection technology on wireless routers equipped with OpenWrt system has good theoretical significance and practical application value for the security protection of intelligent terminal equipment.Based on the systematic study of OpenWrt structure,development process and its development,this paper focuses on key technology research of intrusion detection.Based on the research on intrusion detection processes and intrusion detection methods including anomaly detection,misuse detection,pattern matching,protocol analysis,and machine learning,the ensemble learning algorithm based on machine learning has a good accuracy rate but has limitation of long training time,the feature selection algorithm is introduced to deal with it,and it is verified by the KDDCUP99 dataset,so as to provide a theoretical basis for efficient intrusion detection.Based on this,a FPFE framework for network attack behavior detection is established,which consists of three modules: HDHD,HDHF,and CEL.In the study,a chi-square extreme random tree feature selection algorithm was proposed to be loaded into the HDHF module.HDHF is combined with CEL module to improve the accuracy of intrusion detection and reduce the module running time.Furthermore,the validity of the FPFE framework using the chi-square extreme random tree feature selection algorithm is verified.Finally,an experimental environment is established by using the designed and developed OpenWrt data capture system.The experimental results show that the FPFE framework based on the chi-square extreme random tree feature selection algorithm has an accuracy rate of 96.897% on the OpenWrt dataset and has good practical application value.