An Intelligent AI-Driven Framework for Cyber Attack Detection in Wi-Fi Networks Using Machine Learning and Deep Learning
Year: 2026
Type: conference-paper
Abstract: The emergence of the IEEE 802.11 Wi-Fi networks brings in more complexity in the field of cyber security. There is a need to implement intelligent and adaptive intrusion detection systems (IDSs) to mitigate advanced attacks. Conventional cybersecurity techniques often face challenges in detecting cyberattacks because of the dynamic nature of wireless traffic, high-dimensional features, and constantly changing attack patterns. This paper discusses the state-of-the-art artificial intelligence (AI) methods for intrusion detection in Wi-Fi networks. Different types of AI techniques based on machine learning (ML) and deep learning (DL) methods are reviewed. The existing models, datasets, and strategies are discussed to find out the main problems associated with the feature selection, class imbalance, generalization, and interpretability of AI models. In order to solve these problems, an intelligent and scalable hybrid AI framework is suggested, which consists of feature optimization, SMOTEENN technique, ensemble ML and DL, and Explainable Artificial Intelligence (XAI). In the proposed framework, CNN-DNN models are combined with different ML algorithms, such as Random Forest, XGBoost, and CatBoost. The SHapley Additive exPlanations (SHAP) method is used to improve the interpretability of the model through feature identification.
Authors: Kruti Patel, Vikas Tulshyan, Dr. Rahul Vaghela
Institution: Silver Oak University
Language: English
Subfield: Computer Networks and Communications
Field: Computer Science
Domain: Physical Sciences
Open Access status: gold
OpenAlex ID: W7225150239