This book presents a comprehensive exploration of how artificial intelligence techniques are transforming modern networking systems. It begins with foundational concepts in computer networks, explaining core components such as protocols, transmission media, and network architectures. The introductory chapters bridge traditional networking with machine learning (ML), highlighting how supervised, unsupervised, and reinforcement learning approaches, address challenges. These challenges range from traffic classification, quality-of-service prediction, anomaly detection to dynamic routing. A detailed treatment of deep learning (DL) architectures including CNNs, RNNs, GNNs, autoencoders, GANs, and transformers, demonstrates how complex, high-dimensional network data can be modeled effectively for optimization and security.This book also book introduces lightweight and visual traffic-classification frameworks based on Kolmogorov–Arnold Networks (KAN), including the KAN-Vis model and the RISK-4-Auto architecture for automotive networks. It further presents hybrid deep learning approaches, such as ODENet–LSTM models for botnet detection and an optimized multi-layer intrusion detection system enhanced with genetic algorithms. Each methodology is supported by systematic experimentation and performance evaluation. The concluding chapter outlines future directions in AI-native networking, edge intelligence, federated learning, and self-healing security architectures. This book targets researchers and professional working in this related field as well as graduate students focused on intelligent networking.

This book presents a comprehensive exploration of how artificial intelligence techniques are transforming modern networking systems. It begins with foundational concepts in computer networks, explaining core components such as protocols, transmission media, and network architectures. The introductory chapters bridge traditional networking with machine learning (ML), highlighting how supervised, unsupervised, and reinforcement learning approaches, address challenges. These challenges range from traffic classification, quality-of-service prediction, anomaly detection to dynamic routing. A detailed treatment of deep learning (DL) architectures including CNNs, RNNs, GNNs, autoencoders, GANs, and transformers, demonstrates how complex, high-dimensional network data can be modeled effectively for optimization and security.This book also book introduces lightweight and visual traffic-classification frameworks based on Kolmogorov–Arnold Networks (KAN), including the KAN-Vis model and the RISK-4-Auto architecture for automotive networks. It further presents hybrid deep learning approaches, such as ODENet–LSTM models for botnet detection and an optimized multi-layer intrusion detection system enhanced with genetic algorithms. Each methodology is supported by systematic experimentation and performance evaluation. The concluding chapter outlines future directions in AI-native networking, edge intelligence, federated learning, and self-healing security architectures. This book targets researchers and professional working in this related field as well as graduate students focused on intelligent networking.