Deep Learning-Based Intelligent Anomaly Detection for Cybersecurity in Edge-Cloud Computing Systems

Abstract

The rapid adoption of edge-cloud computing has created highly distributed digital environments in which data, applications, devices, and computational resources operate across heterogeneous edge nodes and centralized cloud infrastructures. Although this architecture improves scalability, responsiveness, and resource utilization, it also introduces complex cybersecurity challenges, including distributed attacks, abnormal network behavior, compromised devices, zero-day threats, and rapidly changing traffic patterns. Conventional rule-based intrusion detection systems often struggle to identify previously unseen or sophisticated anomalies because they depend heavily on predefined signatures and manually engineered features. This paper proposes a deep learning-based intelligent anomaly detection framework for cybersecurity in edge-cloud computing systems. The proposed approach integrates data collection from edge devices, preprocessing, feature extraction, deep representation learning, anomaly classification, and adaptive threat analysis. Deep learning models such as convolutional neural networks, long short-term memory networks, autoencoders, and hybrid architectures can identify spatial, temporal, and behavioral patterns in heterogeneous cybersecurity data. The framework further supports distributed processing at the edge while leveraging cloud resources for large-scale model training and centralized intelligence. The expected outcomes include improved anomaly detection accuracy, reduced detection latency, better identification of unknown attacks, and scalable cybersecurity monitoring. The study demonstrates the potential of intelligent deep learning to strengthen security, resilience, and adaptive threat detection across modern edge-cloud environments.