Mitigating distributed denial of service attacks using attribute subset selection with temporal convolutional...

Area of Science:

• Cybersecurity
• Network Security
• Artificial Intelligence

Background:

• Distributed Denial of Service (DDoS) attacks pose a significant and evolving threat to network infrastructure and services.
• Existing methods struggle with the dynamic patterns and complexity of modern DDoS attacks, necessitating advanced detection techniques.
• Real-time identification and mitigation of DDoS threats are crucial to prevent service disruptions and data breaches.

Purpose of the Study:

• To propose a novel Intelligent Framework for Attack Detection Using Salp Swarm-Based Feature Selection and Deep Learning Architecture (IFAD-SSFSDLA) for real-time DDoS attack detection.
• To enhance the accuracy and efficiency of DDoS attack detection through optimized feature selection and deep learning.
• To provide a robust solution for identifying and mitigating the impact of increasingly sophisticated DDoS attacks.

Main Methods:

• Data preprocessing using min-max normalization for cleaning and structuring raw network traffic data.
• Feature selection employing the Salp Swarm Algorithm (SSA) to identify and retain the most discriminative features for improved model performance.
• Attack classification utilizing the Temporal Convolutional Network (TCN) deep learning architecture.

Main Results:

• The IFAD-SSFSDLA model demonstrated superior performance in detecting DDoS attacks.
• Achieved high accuracy rates of 99.56% on the CIC-IDS-2017 dataset and 99.65% on the Edge-IIoT dataset.
• Outperformed existing techniques in DDoS attack detection accuracy across multiple datasets.

Conclusions:

• The proposed IFAD-SSFSDLA model offers an effective and accurate solution for real-time DDoS attack detection.
• The integration of Salp Swarm Algorithm for feature selection and Temporal Convolutional Network for classification significantly enhances detection capabilities.
• This framework provides a valuable advancement in cybersecurity for combating prevalent and evolving network threats.