Multiple Immune-based Approaches for Network Traffic Analysis
Inadyuti Dutt, Samarjeet Borah, Indra Kanta Maitra · Procedia Computer Science · 2020
Internet and intranet facilities face security related issues due to malicious attacks in the network and unwanted access by unknown or sometimes known users. The network for these facilities becomes vulnerable and therefore, the network needs to be analysed for finding intrusive activities. In this paper, multiple immune-based approaches have been proposed for analyzing the network. The main intention of our work is to detect such vulnerabilities by capturing the traffic and analyzing both the header and the payload portions of the network packet. This paper tries to employ a scheme that would consider both the header and payload-portions of a network packet. The standard KDD’99 data set has been used for analysis and features are extracted using the well-known method called PCA. Then the selected features are supplied to the various immune-based algorithms for detecting anomalies in the network. The immune-based algorithms AIRS-inspired, Clonal Selection-inspired and Immunity structure inspired are implemented and the results show that the AIRS-inspired approaches perform much better than the other immune-based approaches for all the tcp, icmp and udp packets.