Design of Anomaly-Based Intrusion Detection System Using Fog Computing for IoT Network

With increase in the demand for Internet of Things (IoT)-based services, the capability to detect anomalies such as malicious control, spying and other threats within IoT-based network has become a major issue. Traditional Intrusion Detection Systems (IDSs) cannot be used in typical IoT-based networ...

Full description

Saved in:
Bibliographic Details
Published in:Automatic control and computer sciences Vol. 55; no. 2; pp. 137 - 147
Main Authors: Prabhat Kumar, Gupta, Govind P., Tripathi, Rakesh
Format: Journal Article
Language:English
Published: Moscow Pleiades Publishing 01.03.2021
Springer Nature B.V
Subjects:
Accuracy
Algorithms
Anomalies
Cloud computing
Computer Science
Control Structures and Microprogramming
Decision trees
Espionage
False alarms
Internet of Things
Intrusion detection systems
Machine learning
Model testing
fog computing
intrusion detection system
anomaly detection
Internet of Things
feature selection
ISSN:0146-4116, 1558-108X
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:With increase in the demand for Internet of Things (IoT)-based services, the capability to detect anomalies such as malicious control, spying and other threats within IoT-based network has become a major issue. Traditional Intrusion Detection Systems (IDSs) cannot be used in typical IoT-based network due to various constraints in terms of battery life, memory capacity and computational capability. In order to address these issues, various IDSs have been proposed in literature. However, most of the IDSs face problem of high false alarm rate and low accuracy in anomaly detection process. In this paper, we have proposed a anomaly-based intrusion detection system by decentralizing the existing cloud based security architecture to local fog nodes. In order to evaluate the effectiveness of the proposed model various machine learning algorithms such as Random Forest, K-Nearest Neighbor and Decision Tree are used. Performance of our proposed model is tested using actual IoT-based dataset. The evaluation of the underlying approach outperforms in high detection accuracy and low false alarm rate using Random Forest algorithm.
Bibliography:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:0146-4116
1558-108X
DOI:10.3103/S0146411621020085