REAL-TIME ANOMALY DETECTION IN DISTRIBUTED SYSTEMS USING FEDERATED LEARNING TECHNIQUES
Abstract
Anomaly detection in distributed systems is crucial for ensuring system reliability, fault tolerance, and security. Traditional centralized learning models often suffer from data privacy concerns, communication overhead, and latency. This paper explores the integration of Federated Learning (FL) into real-time anomaly detection mechanisms within distributed environments, leveraging edge computing nodes to collaboratively train anomaly detection models without data centralization. We propose a framework that integrates federated averaging and LSTM-based models for sequential anomaly detection, which preserves privacy and reduces computational delays. Simulated experiments on benchmark datasets such as KDDCup and NSL-KDD demonstrate the efficacy of our method in identifying anomalies with high accuracy and low latency