Hybrid Intrusion Detector using Deep Learning Technique
G. Kanagaraj, T. Primya, G. Subashini, V. Senthilkumar, S. Gomathi · 2021 International Conference on Advancements in Electrical, Electronics, Communication, Computing and Automation (ICAECA) · 2021
To shelter a computer network infrastructure, to avoid illegal changes and mistreatment in system, besides hacking network security is to defend computer network. To realize network security a method called firewall is used. Depends on set of rules incoming and outgoing network traffics are controlled, it may be based on hardware or software. By using already available system vulnerabilities process of making the system unstable, process of manipulating information, accessing unauthorized system by careful actions is defined as intrusion or threat called network attack. Unwanted activity of taking up network resources meant for others, vulnerability of network security is network intrusion. In computer networks, Intrusion Detection system (IDS) / Intrusion Prevention System (IPS) are required as prior condition. To monitor network and harmful activities in system a device or software application called IDS/IPS is used. If it is used in network side called as Network-Based IDS/IPS. Bulky numbers of attacks emerging day after day in network. It is unable to detect new attacks by current network intrusion detection due to assortment of reasons like amount of data, miscellany of data, low frequency attacks and more. This paper reduces the human effort obligatory to enlarge the model like data pre processing, feature selection. Found Deep learning as solution to over listed problems. It depends on promising deep learning model and integrating it with software and tested across an assortment of deep learning model and chooses the best one with additional accuracy. Functionality includes intrusion detection, malware detection, and traffic analysis. Model is trained on NSL dataset which found benchmark all over the world. It helps to make the network more secure and free of attacks.