Protocols comparison for real-time data streaming from IoT devices to a cloud-based solution

Abstract

The proper detection and prevention of malfunctions are crucial in mitigating maintenance costs and equipment replacements for agricultural vehicles, ultimately reducing the expenses associated with crop cultivation. Predictive analytics for agriculture vehicles leverage machine learning and sensor data to anticipate equipment faults, optimize maintenance schedules, and enhance operational efficiency in the farming industry. It heavily relies on real-time data transmission to continuously monitor equipment performance, enabling timely identification of potential issues and preemptive maintenance actions to prevent costly breakdowns and downtime. This paper employs a qualitative analysis approach utilizing the Architecture Tradeoff Analysis Method to evaluate and select an optimal data protocol from a set of candidates, including SOAP, HTTP, REST, CoAP, WebSocket, XMPP, MQTT, and AMQP. The analysis considers sensitivity points, tradeoff factors, risks, and quality attribute scenarios relevant to the usage scenarios. The findings indicate that MQTT is the preferred protocol for real-time data streaming in resourceconstrained environments, contingent upon a reliable connection.