Title: IBBO-LSSVM-based network anomaly intrusion detection

Authors: Peng Zhou; Wen-Kuang Chou

Addresses: School of International Education, Huanghuai University, Henan, 463000, China; Laboratory of Behavior and Security based on Deep Learning and Big Data, Henan, 463000, China ' Department of Computer Science and Information Engineering, Providence University, Taichung, 43301, Taiwan

Abstract: Owing to the variety and complexity of network intrusion, the traditional network anomaly intrusion detection model cannot accurately classify and identify the abnormal intrusion behaviour of the network, resulting in poor performance when detecting the network anomaly intrusion. In order to improve the performance of network intrusion detection, we propose a novel network anomaly intrusion detection method, by means of IBBO-LSSVM. In this paper, the least squares support vector machine is applied to model and analyse the network abnormal intrusion detection, which can capture the relationship between network anomaly intrusion types and its corresponding features. Then, an improved biogeography-based approach is applied to optimise the parameters of the network intrusion detection model. Finally, the model is simulated and evaluated on a standard network anomaly intrusion test database. The accuracy of the network anomaly intrusion detection for the proposed method is higher than 90%, demonstrating that the proposed approach is superior to the traditional methods.

Keywords: abnormal network behaviour; intrusion detection; modelling and analysis; improved biogeography-based optimisation; IBBO; support vector machine; SVM.

DOI: 10.1504/IJES.2018.10017165

International Journal of Embedded Systems, 2019 Vol.11 No.3, pp.352 - 362

Received: 01 May 2018
Accepted: 12 Jun 2018

Published online: 23 Apr 2019 *

Full-text access for editors Access for subscribers Purchase this article Comment on this article