Next-Generation Cyber Defense Using Threat Intelligence Fusion Across Cloud, Edge, and IoT Environments

Abstract

The exponential expansion of interconnected heterogeneous digital ecosystems—spanning hyper-scale cloud data centers, multi-access edge computing (MEC) gateways, and resource-constrained Internet of Things (IoT) deployments—has fundamentally invalidated traditional perimeter based cyber defense paradigms. Modern Advanced Persistent Threats (APTs) and multi-stage cyber attacks exploit the silos separating cloud, edge, and IoT layers to execute sophisticated, low-and-slow campaigns that evade isolated detection mechanisms. This comprehensive review paper investigates the state-of-the-art in next-generation cyber defense mechanisms powered by multi-source Cyber Threat Intelligence (CTI) fusion across heterogeneous environments. We perform an exhaustive synthesis of recent literature (2019 2026), detailing data integration models, threat correlation architectures, machine learning frameworks, and automated response mechanisms. Furthermore, we analyze the integration of Dynamic Graph Neural Networks (DGNNs) and Privacy-Preserving Federated Learning (FL) to enable real time threat intelligence sharing without compromising data privacy or bandwidth constraints. Based on rigorous structural comparison, we evaluate current methodologies across critical technical metrics including detection latency, scalability, noise sensitivity, and adversarial robustness. Finally, we articulate pressing open research challenges and present strategic future directions toward autonomous, zero-trust cyber defense ecosystems.

KEYWORDS: Cyber Threat Intelligence (CTI), Data Fusion, Dynamic Graph Neural Networks, Federated Learning, Multi-Access Edge Computing, Cloud Security, IoT Defense, Zero Trust Architecture.