Related Experiment Video
Updated: May 21, 2025
03:31
Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications
Published on: December 15, 2023
448
Securing IoT Networks Against DDoS Attacks: A Hybrid Deep Learning Approach.
Noor Ul Ain1, Muhammad Sardaraz1, Muhammad Tahir1
1Department of Computer Science, COMSATS University Islamabad, Attock Campus, Attock 43600, Pakistan.
Sensors (Basel, Switzerland)
|March 17, 2025
Summary
This study introduces a hybrid deep learning model for detecting Distributed Denial-of-Service (DDoS) attacks in Internet of Things (IoT) networks. The novel approach significantly improves detection accuracy, outperforming existing methods on the CICIOT2023 dataset.
Area of Science:
- Cybersecurity
- Network Security
- Artificial Intelligence
Background:
- The Internet of Things (IoT) is rapidly expanding, leading to increased network complexity and vulnerability.
- Growing interconnectivity in IoT networks presents significant security challenges, particularly concerning Distributed Denial-of-Service (DDoS) attacks.
- Effective detection of sophisticated DDoS attacks is crucial for maintaining the integrity and availability of IoT systems.
Purpose of the Study:
- To apply and evaluate deep learning techniques for enhanced DDoS attack detection in IoT environments.
- To propose and validate a novel hybrid deep learning model for improved IoT network security against DDoS threats.
- To assess the performance of various deep learning architectures in identifying complex attack patterns.
Main Methods:
- Assessed performance of Latent Autoencoders, LSTM Autoencoders, and Convolutional Neural Networks (CNNs) for DDoS detection.
- Developed a hybrid model integrating CNNs (feature extraction), Long Short-Term Memory (LSTM) networks (temporal patterns), and Autoencoders (dimensionality reduction).
- Utilized the CICIOT2023 dataset for experimental evaluation and performance comparison.
Main Results:
- The proposed hybrid model achieved high accuracy, with 96.78% training accuracy and 96.60% validation accuracy.
- Experimental results demonstrated that the hybrid model significantly outperformed individual deep learning models in detecting DDoS attacks.
- The model showed efficiency in addressing complex attack patterns prevalent in modern IoT networks.
Conclusions:
- The developed hybrid deep learning model offers a robust solution for DDoS attack detection in IoT networks.
- Further research is needed to address limitations in detecting rare attack types and data imbalance issues.
- Enhancing DDoS detection capabilities requires continued focus on advanced deep learning techniques and comprehensive datasets.
