Intrusion detection in machine learning based E-shaped structure with algorithms, strategies and applications in...

Area of Science:

• Computer Science
• Network Security
• Wireless Communication

Background:

• Distributed sensor networks are crucial for modern applications like cloud computing and the Internet of Things (IoT).
• Existing intrusion detection systems (IDS) struggle with anomaly detection and identifying zero-day attacks in diverse IoT data streams.
• Current signature-based IDS are insufficient for the evolving threat landscape in IoT environments.

Purpose of the Study:

• To introduce a novel Filtered Deep Learning Model for Intrusion Detection with a Data Communication Approach.
• To enhance the detection of zero-day attacks in IoT networks by addressing challenges with varied data sources.
• To improve the accuracy and efficiency of intrusion detection systems in distributed sensor networks.

Main Methods:

• The proposed model involves five stages: Sensor Network Initialization, Cluster Formation and Head Selection, Connectivity, Attack Detection, and Data Broker.
• A filtered deep learning approach is employed for anomaly detection within the sensor network.
• An E-shaped patch antenna was designed, fabricated, and tested for wide-band wireless communication.

Main Results:

• The Filtered Deep Learning Model achieved a superior accuracy of 96.12% in intrusion detection compared to existing Deep Learning Neural Net and Artificial Neural Network models.
• Experimental results demonstrated the model's effectiveness in identifying intrusions in IoT networks.
• The developed E-shaped patch antenna exhibited a wide bandwidth, resonating at 7.5 and 8.5 GHz, with detailed performance characteristics presented.

Conclusions:

• The proposed filtered deep learning model significantly improves intrusion detection accuracy in IoT networks, outperforming traditional methods.
• The study highlights the potential of advanced deep learning techniques for securing distributed sensor networks against sophisticated cyber threats.
• The novel E-shaped patch antenna design offers a promising solution for efficient wide-band wireless communication in IoT applications.