Performance Analysis-Based Parameter Tuning of Clonal Selection Algorithm for Anomaly Detection

Abstract

Intrusion detection has become paramount in the field of network security owing to the fact that network data are being compromised on a daily basis. To this effect, several algorithms have been made available to detect intrusion in the network environment. The Clonal Selection Algorithm (CSA) is one of such algorithms for intrusion detection. Often times, the detection capability of this algorithm are limited by incorrect settings of the parameters involved. Thus, tuning some of the parameters involved in CSA is sacrosanct in determining the performance analysis of the algorithm. Hence, this paper is aimed at tuning the parameters of CSA and analysing its performance for anomaly-based intrusion detection using KDDcup’99 dataset as benchmark for the evaluation. The findings showed that CSA is a good intrusion detection algorithm.