Machine Learning and Optimization Techniques for Cyberattack Prevention in Wireless Networks
Usharani Bhimavarapu · 2025
Cybersecurity in modern systems, particularly in the context of space systems and network infrastructure, faces increasing challenges due to the sophistication and variety of attacks. To effectively secure such systems, it is crucial to implement robust detection and monitoring methods that can identify malicious activity in real-time. This study explores the application of machine learning techniques, particularly random forest and feature selection methods like particle swarm optimization (PSO), to enhance cybersecurity measures. The AWID2 dataset, sourced from IEEE 802.11 wireless networks, was utilized to simulate real-world network conditions and different types of attacks, including impersonation, injection, and flooding. The research focused on pre-processing the dataset through normalization and scaling, followed by feature extraction using principal component analysis (PCA) and feature selection using PSO. The performance of the system was evaluated by training a random forest classifier, demonstrating its ability to effectively differentiate between normal and attack traffic. The findings suggest that the combination of these techniques can significantly improve the accuracy and efficiency of attack detection systems, offering insights into how machine learning can be integrated into cybersecurity strategies for space and network infrastructures.