ADAPTIVE MODEL OF CYBERTHREAT DETECTION IN IOT SUBSYSTEMS DATE CENTERS

Abstract

The integration of Internet of Things (IoT) technologies, edge/fog computing, and cloud services into modern data centers creates prerequisites for improving monitoring efficiency, infrastructure management automation, continuous telemetry data collection, and adaptive control of cooling systems, power consumption, and network resources within data centers. Alongside these advantages, the number of cyber threats is also increasing due to the expansion of the attack surface, the use of heterogeneous IoT devices, and the complexity of securing distributed infrastructures. This paper investigates the features of implementing IoT components in cloud environments and data centers while considering modern information security threats. Particular attention is paid to DDoS attacks, spoofing attacks, and the compromise of infrastructure nodes. An adaptive cyber threat detection model is proposed, combining statistical network traffic analysis, behavioral analysis of nodes, and graph-based representation of interactions between system components. To assess the state of the infrastructure, an integrated risk indicator is used, taking into account traffic intensity, source entropy, behavioral profile deviation, and interaction graph parameters. The paper also proposes the architecture of an adaptive cyber threat detection system for an IoT-oriented data center, including the IoT layer, the edge/fog preprocessing layer, the analytical cloud layer, the graph correlation module, and the automated cyber incident response layer.