Measuring and Adapting MQTT in Cellular Networks for Collaborative Smart Farming

Abstract

In distributed Smart Farming applications, a reliable communication is often crucial, in particular during collaborating operations. In rural areas, the network connectivity via Public Land Mobile Networks is, however, not always sufficient. Network disruptions may occur hindering reliable transmissions. For the delivery of periodic sensor data streams of agricultural machines, several modern communication frameworks adopt the Message Queue Telemetry Transport (MQTT). Although MQTT is built on top of TCP, a reliable delivery, even of important data, cannot generally be guaranteed in rural environments. Due to handovers and spatial dead zones, this is in particular the case when clients are mobile.,,,,This paper presents a case-study-based performance evaluation of MQTT regarding a reliable data transmission in real-world scenarios with mobile clients. Therefore, we investigate both links, publisher to broker and broker to subscriber, separately. By doing so, we show that both links significantly benefit from a suitable parameterization of MQTT. Moreover, for many data streams in the considered scenario, the default FIFO queuing strategy of MQTT is not always the most suitable approach to cope with network disruptions. Hence, we implemented LIFO queuing as well as a novel hybrid approach in MQTT as a proof-of-concept and evaluated its impact.