Context-aware enhancement of buffer utilisation in MQTT-based IoT communication

This paper explores Internet of Things (IoT) communication and the crucial role of buffer utilisation in enhancing the message queuing telemetry transport (MQTT) protocol. Unlike traditional approaches that highlight the benefits of buffering, this study takes a distinctive perspective by addressing the limitations of buffer usage in IoT communication. The research introduces a context-aware strategy to employ buffers judiciously, considering specific data requirements and network connectivity scenarios. Experimental setups involving Node-RED, Wireshark, and prominent MQTT brokers (HiveMQ, EMQx, and Mosquitto) facilitate comprehensive performance analysis. Parameters such as connection setup time, subscription time, mean delay, mean jitter, and bandwidth consumption are evaluated. The findings consistently demonstrate performance degradation after broker reconnection, with more prolonged network failures leading to more significant degradation. The study emphasises the importance of using buffers wisely, especially for prioritising critical messages, to ensure reliability and enhance overall MQTT network performance.