An Integrated Approach to Network Intrusion Detection and Prevention

Abstract

At present, with the expansion of size of the internet, security plays a crucial role in computer networks. Also with the advancement of Internet of things, earlier technology like firewall, authentication and encryption are not effective in ensuring the complete security. This has lead to the development of Intrusion Detection Systems (IDS) which monitors the events in computer networks to recognize the threats that violates computer security. With the help of various machine learning algorithms we have carried out the implementation of IDS. Machine learning technique increases the accuracy of anomaly detection in real-time scenario. This work focuses on K-Nearest Neighbor (KNN) classifier and Support Vector Machine (SVM), which classify the program behavior as intrusive or not. To prevent DoS (Denial-of-Service) attacks, a new method is implemented in this paper. The KNN classified data which provides malicious IP address are blocked in routers through Standard Access-list.