Adaptive Network Monitoring Framework for IoT Communication Protocols

Abstract

The fast development of heterogeneous Internet of Things (IoT) ecosystems has turned real-time network monitoring into a necessary element to be taken into consideration to guarantee reliability and security. In this paper, a flexible monitoring system is described that uses passive metering methods to examine communication protocols like MQTT, CoAP and AMQP in large internet of things implementation. The framework has applied flow level metadata, time based correlation, and profiling of devices to identify anomalies without probing. Experimental testing of 1,200 IoT devices incorporates 92 percent accuracy in detecting the aberrant behaviour and protocol abuse. The presented solution offers a protocol-agnostic network intelligence base and enables the following-generation IoT traffic auditing and intrusion prevention system.