Intelligent Anomaly Detection in Cloud-Native Distributed Systems: A Deep Learning–Driven Framework for Scalable and Adaptive Infrastructure Monitoring
International Journal of Scientific Research in Science Engineering and Technology · 01.05.2023 · 890
DOI: 10.32628/ijsrset23107112 ↗Kurz gefasst
Fachartikel in der Zeitschrift International Journal of Scientific Research in Science Engineering and Technology, online erschienen am 22.05.2023. Verfasst von Madhava Rao Thota Madhava Rao Thota. Die Arbeit nennt 23 Literaturangaben.
Abstract (Originalsprache)
Modern distributed databases and cloud-native infrastructures generate vast volumes of heterogeneous data, including logs, metrics, and traces, produced continuously by microservices, containers, and orchestration platforms operating at scale. Detecting anomalies within such systems is critical for ensuring reliability, security, and performance, as even minor deviations can cascade into large-scale service disruptions or security incidents. Traditional rule-based and statistical methods, while effective in simpler environments, struggle to scale with the complexity, high dimensionality, and dynamic behavior of modern cloud ecosystems, where system states evolve rapidly and dependencies are often non-linear and opaque. This paper presents a comprehensive study of deep learning–driven anomaly detection techniques tailored for distributed systems, emphasizing their ability to automatically learn hierarchical feature representations and temporal dependencies directly from raw data. By synthesizing architectural workflows, taxonomy models, and system-level implementations, we highlight how deep learning models such as Long Short-Term Memory (LSTM) networks, autoencoders, and Generative Adversarial Networks (GANs) significantly improve anomaly detection accuracy, robustness, and adaptability across diverse operational scenarios. Furthermore, the study explores the integration of these models into real-time monitoring pipelines, discusses challenges such as concept drift, data imbalance, and interpretability, and identifies future research directions focused on scalable, explainable, and resource-efficient anomaly detection in increasingly decentralized and cloud-native environments.
Zitieren
APA
Thota, M. R. T. M. R. (2023). Intelligent Anomaly Detection in Cloud-Native Distributed Systems: A Deep Learning–Driven Framework for Scalable and Adaptive Infrastructure Monitoring. International Journal of Scientific Research in Science Engineering and Technology, 890. https://doi.org/10.32628/ijsrset23107112
Vancouver
Thota MRTMR. Intelligent Anomaly Detection in Cloud-Native Distributed Systems: A Deep Learning–Driven Framework for Scalable and Adaptive Infrastructure Monitoring. International Journal of Scientific Research in Science Engineering and Technology. 2023:890. doi:10.32628/ijsrset23107112
BibTeX
@article{Thota2023,
author = {Thota, Madhava Rao Thota Madhava Rao},
title = {Intelligent Anomaly Detection in Cloud-Native Distributed Systems: A Deep Learning–Driven Framework for Scalable and Adaptive Infrastructure Monitoring},
journal = {International Journal of Scientific Research in Science Engineering and Technology},
year = {2023},
pages = {890},
doi = {10.32628/ijsrset23107112}
}
EUVIMED