An improved deep autoencoder-based network intrusion detection system with enhanced performance
by Bidyapati Thiyam; Shouvik Dey
International Journal of Internet Technology and Secured Transactions (IJITST), Vol. 13, No. 3, 2024

Abstract: The evolutions in information and communication technology (ICT) devices have led to the rise in cyber-attacks in network systems. Ensuring cybersecurity in these devices is one of the most critical and complex issues. Besides, with technological improvement, hackers have been introducing complex and malicious malware attacks into the network system, making intrusion detection a complicated task. However, various conventional IDS face severe challenges to discern and counteract such risks. Statistical research and an enhanced autoencoder classifier for the network intrusion detection system are introduced in this paper (NIDS). This work uses an optimised autoencoder feature extraction process to extract correlated features to simplify the classification process. The proposed method is estimated using the NSL-KDD, UNSW-NB15, and CIC-DDoS2019 datasets, and its performance is compared to that of current machine learning algorithms. The results indicate that the proposed AE-NIDS outperforms shallow machine learning classifiers regarding detection accuracy and false positive rate.

Online publication date: Thu, 15-Feb-2024

The full text of this article is only available to individual subscribers or to users at subscribing institutions.

 
Existing subscribers:
Go to Inderscience Online Journals to access the Full Text of this article.

Pay per view:
If you are not a subscriber and you just want to read the full contents of this article, buy online access here.

Complimentary Subscribers, Editors or Members of the Editorial Board of the International Journal of Internet Technology and Secured Transactions (IJITST):
Login with your Inderscience username and password:

Username:        Password:          Forgotten your password?

Want to subscribe?
A subscription gives you complete access to all articles in the current issue, as well as to all articles in the previous three years (where applicable). See our Orders page to subscribe.

If you still need assistance, please email subs@inderscience.com