Feature engineering and deep learning-based intrusion detection framework for securing edge IoT

Devices belonging to the realm of edge Internet of Things (IoT) are becoming highly susceptible to intrusion attacks. The large-scale development in edge IoT, ease of availability, and affordability have drastically increased its usage in the real world. The business market revolves around producing...

Full description

Saved in:
Bibliographic Details
Published in:The Journal of supercomputing Vol. 78; no. 6; pp. 8852 - 8866
Main Authors: Nasir, Muneeba, Javed, Abdul Rehman, Tariq, Muhammad Adnan, Asim, Muhammad, Baker, Thar
Format: Journal Article
Language:English
Published: New York Springer US 01.04.2022
Springer Nature B.V
Subjects:
Artificial neural networks
Compilers
Computer Science
Data Mining for IoT in Mobile Edge computing
Deep learning
Internet of Things
Interpreters
Intrusion detection systems
Machine learning
Processor Architectures
Programming Languages
Deep learning
Network security
Edge IoT
Intrusion detection system
Machine learning
ISSN:0920-8542, 1573-0484
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Devices belonging to the realm of edge Internet of Things (IoT) are becoming highly susceptible to intrusion attacks. The large-scale development in edge IoT, ease of availability, and affordability have drastically increased its usage in the real world. The business market revolves around producing better, innovative, and appealing products every day. However, security is often left unchecked to achieve these standards. Therefore, vulnerabilities present in these devices make them susceptible to various intrusion attacks. We devised a model named DF-IDS for detecting intrusions in IoT traffic. DF-IDS consists of two main phases: In the 1 st phase, it comparatively selects the best features from the feature matrix using SpiderMonkey (SM), principle component analysis (PCA), information gain (IG), and correlation attribute evaluation (CAE). In the 2 nd phase, these features along with assigned labels are used to train a deep neural network for intrusion detection. DF-IDS achieves an accuracy of 99.23% with an F1-score of 99.27%. It shows improvement not only in accuracy but also in F1 score as compared to the other comparative models and existing studies.
Bibliography:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:0920-8542
1573-0484
DOI:10.1007/s11227-021-04250-0