Reducing Energy Consumption in Continuous Cloud Threat Monitoring Through Adaptive Sampling

Continuous threat monitoring of cloud-based systems is critical for recognizing potential cyber attacks; identifying unusual behavior; and safeguarding confidential data from dynamic, evolving cloud-based systems. However, ongoing gathering and analysis of security related telemetry can require significant amounts of computing power and energy, especially in larger-scale systems with large volumes of logging; network-event; authentication; and system-metrics data. This study presents a dynamic sampling methodology to provide both energy savings and adequate threat detection. The dynamic sampling framework adjusts the monitoring rate based on workload level; previous threat history; system sensitivity; and current indicators of possible threats. When the risk level is lower, the framework decreases the sampling rate to eliminate unnecessary data processing, storage, and computation. When anomalies occur or when the risk level increases, the monitoring system increases the sampling rate to enhance visibility and enable quick identification of incidents. Risk-based decision making, feedback-driven adaptation, and energy-aware resource allocation are used in the development of the framework to balance security with operational efficiency. Simulations of cloud workloads and representative security event data sets could be used to evaluate the dynamic sampling framework by measuring energy consumption; detection accuracy; false-positive rates; response times; and monitoring overhead. This research establishes a theoretical basis for sustainable cloud security monitoring and provides guidance on implementation issues such as sampling fairness; gaps in visibility; and resistance to adversarial manipulation.

Cloud security, adaptive sampling, energy efficiency, threat monitoring, cloud computing, anomaly detection, security telemetry, sustainable computing, resource optimization, cybersecurity.