Network intrusion detection using self-recurrent wavelet neural network with multidimensional radial wavelons

Abstract

In this paper we report a novel application-based model as a suitable alternative for the classification and identification of attacks on a computer network, and thus guarantee its safety from HTTP protocol-based malicious commands. The proposed model is built on a self-recurrent neural network architecture based on wavelets with multidimensional radial wavelons, and is therefore suited to work online by analyzing non-linear patterns in real time to self-adjust to changes in its input environment. Six different neural network based systems have been modeled and simulated for comparison purposes in terms of overall performance, namely, a feed-forward neural network, an Elman network, a fully connected recurrent neural network, a recurrent neural network based on wavelets, a self-recurrent wavelet network and the proposed self-recurrent wavelet network with multidimensional radial wavelons. Within the models studied, this paper presents two recurrent architectures which use wavelet functions in their functionality in very distinct ways. The results confirm that recurrent architectures using wavelets obtain superior performance than their peers, in terms not only of the identification and classification of attacks, but also the speed of convergence.