Cloud-Based Real-Time Anomaly Detection in Network Traffic
K Dinesh Kumar, D. K. Niranjan, P Praneeth Reddy, P Himanshu, K Sai Raju, Praneeth Goud · 2025
In the era of increasing cyber threats and growing network complexity, real-time anomaly detection plays a pivotal role in ensuring network security. This paper presents a scalable, cloud-based framework for detecting anomalies in network traffic using machine learning models which are trained on the UNSW-NB15 dataset. The proposed system is deployed entirely on Google Cloud Platform (GCP), leveraging Cloud Functions for model inference, BigQuery for storage and analytics, and Streamlit for an interactive user interface and monitoring dashboard. The architecture enables real-time processing and alerting through automated email notifications upon anomaly detection. Comprehensive experiments demonstrate the system’s ability to identify suspicious network behavior with high accuracy and low latency, validating its effectiveness for practical deployment in modern cybersecurity infrastructures. This research highlights the benefits of combining cloud-native tools with data-driven anomaly detection to build a responsive and extensible security monitoring solution.