Exploring Privacy and Security in Distributed Cloud Computing: A Survey on Federated Learning and Advanced Techniques

Abstract

Distributed cloud computing in India has expanded rapidly across sectors like
finance, healthcare, and government, leading to increased concerns about data privacy and cyber
threats. Reports indicate that a significant number of organizations have experienced cloudrelated
security incidents, highlighting the need for stronger protection mechanisms. Traditional
manual security systems rely on human monitoring, rule-based controls, and predefined
thresholds, making them slow, error-prone, and unable to scale effectively in complex distributed
environments. They often fail to detect sophisticated attacks and provide delayed responses. The
proposed system enhances security by integrating CRN-GRIM with federated learning. CRNGRIM
enables real-time anomaly detection using deep learning, while federated learning ensures
data privacy by keeping sensitive information on local nodes and sharing only model updates.
This approach improves detection accuracy compared to PAC and adapts to evolving threats
automatically. Overall, machine learning-driven automation makes distributed cloud security
more efficient, scalable, and privacy-preserving, especially for sensitive applications in India.