A Coronavirus Herd Immunity Optimizer For Intrusion Detection System

Intrusion Detection System (IDS) is considered as one of the essential components of a secure network. Due to the high number of network packet features, one of the major problems of IDS is false intrusion alerts and low intrusion detection rates. Feature selection removes all redundant or irrelevant features among the various features of network packets. For this reason, it plays a pivotal role in overcoming the mentioned problems and can improve the accuracy of intrusion detection system. In this paper, a new human-inspired optimization algorithm called coronavirus herd immunity optimizer (CHIO) is proposed for feature selection in IDS. CHIO is able to select the optimal subset of features from numerous features without affecting system performance. In order to select the feature, two types of classifiers, K-Nearest Neighbor (KNN) and Artificial Neural Network (ANN), are used to obtain the accuracy of intrusion detection. In addition, the ANN classifier is trained with the classic Gradient Descent (GD) method as well as the two intelligent methods Artificial Bee Colony (ABC) and Harmony Search (HS). In order to demonstrate the performance, our method is tested on 20% of NSL-KDD and its results are reported and compared to other studies. The proposed method has been able to achieve better performance in terms of intrusion detection accuracy and number of features compared to similar works.