Developing an Intelligent System for Efficient Botnet Detection in IoT Environment DOI
Ramesh Singh Rawat, Manoj Diwakar, Umang Garg

et al.

International Journal of Mathematical Engineering and Management Sciences, Journal Year: 2025, Volume and Issue: 10(2), P. 537 - 553

Published: Feb. 7, 2025

Smart technological instruments and Internet of Things (IoT) systems are now targeted by network attacks because their widespread rising use. Attackers can take over IoT devices via botnets, pre-configured attack vectors, use them to do harmful actions. Thus, effective machine learning is required solve these security issues. Additionally, deep with the necessary elements advised defend from threats. In order achieve proper detection hacks in future, relevant datasets must be used. The device's operation could occasionally delayed. sample dataset well structured for training model validating suggested create best protection system feasible detecting cyber risks. This paper focused on analyzing botnet traffic an environment using classifiers: Decision tree classifier, Naïve Bayes, K nearest neighbor, Convolution neural network, Recurrent Random Forest. We calculated each algorithm's Accuracy, True Positive, False Negative, Precision, Recall. obtained impressive results CNN, LSTM RNN classifiers. have also achieved a high rate.

Language: Английский

Coarse and fine feature selection for Network Intrusion Detection Systems (IDS) in IoT networks DOI
Mohammed Sayeeduddin Habeeb,

Tummala Ranga Babu

Transactions on Emerging Telecommunications Technologies, Journal Year: 2024, Volume and Issue: 35(4)

Published: March 19, 2024

Abstract Network Intrusion Detection Systems (NIDSs) are important in safeguarding networks from known and unknown attacks. Many research efforts have recently been made to create NIDS systems based on Machine Learning (ML) methods, addressing a significant challenge designing standard the lack of standardized feature sets dataset. Given recent development Internet Things (IoT) wireless communication, our proposed method introduces novel solution enhance intrusion detection systems. This selection is carried out two stages, coarse fine selection. In first stage process, we conduct correlation analysis identify relationships within set. The second employs using Whale Optimization Algorithm (WOA) with Genetic hybridization (CFWOAGA). fitness each selected assessed K‐Nearest Neighbors (KNN) algorithm. work integrate WOA hybrid GA extend search space avoid local optima problems via crossover mutation operations. These features critical for detecting any intrusion, use an ML classifier whether there attack or normal network evaluate performance classifier. We BoT‐IoT 2020 dataset while limiting 32 reduced computational complexity, these upon considerations system optimization efficiency, making balance between efficiency model performance. experimental findings show better accuracy compared technique drop False Alarm Rate (FAR). conclusion, CFWOA achieved 98.9%, updated version genetic algorithm demonstrated further improvement at 99.5%. Notably, was substantial FAR method.

Language: Английский

Citations

6
Developing an Intelligent System for Efficient Botnet Detection in IoT Environment DOI
Ramesh Singh Rawat, Manoj Diwakar, Umang Garg

et al.

International Journal of Mathematical Engineering and Management Sciences, Journal Year: 2025, Volume and Issue: 10(2), P. 537 - 553

Published: Feb. 7, 2025

Smart technological instruments and Internet of Things (IoT) systems are now targeted by network attacks because their widespread rising use. Attackers can take over IoT devices via botnets, pre-configured attack vectors, use them to do harmful actions. Thus, effective machine learning is required solve these security issues. Additionally, deep with the necessary elements advised defend from threats. In order achieve proper detection hacks in future, relevant datasets must be used. The device's operation could occasionally delayed. sample dataset well structured for training model validating suggested create best protection system feasible detecting cyber risks. This paper focused on analyzing botnet traffic an environment using classifiers: Decision tree classifier, Naïve Bayes, K nearest neighbor, Convolution neural network, Recurrent Random Forest. We calculated each algorithm's Accuracy, True Positive, False Negative, Precision, Recall. obtained impressive results CNN, LSTM RNN classifiers. have also achieved a high rate.

Language: Английский

Citations

0