Enhanced Security In Matter-Enabled Iot Networks Through Anomaly Detection
Manjit Kumar, Kapil Dev Sharma · 2025
The adoption of Matter-enabled Internet of Things (IoT) networks has significantly improved interoperability and security in smart environments. However, despite these advancements, mesh networks within the Matter protocol remain vulnerable to security threats, particularly packet flooding attacks. These attacks generate excessive network traffic, causing packet loss, delays, and overall performance degradation. While Matter includes robust security features such as end-to-end encryption, authentication, and replay protection, these measures are insufficient for detecting flooding attacks in real time.This research proposes an intelligent anomaly detection framework that utilizes telemetry data to identify devices contributing to harmful traffic. By employing the Isolation Forest algorithm, the framework accurately distinguishes abnormal network behaviors—such as increased retry attempts and prolonged message delivery times—from normal operations. Experimental results demonstrate its ability to detect malicious activity with high accuracy and minimal false positives.The study underscores the importance of real-time anomaly detection in protecting Matter IoT networks from flooding attacks. The proposed solution is both scalable and resource-efficient, making it well-suited for deployment in resource-constrained IoT environments.