MQTT Systems: A Survey

Abstract: In today's era, billions of intelligent devices populate Internet of Things (IoT) environments worldwide, spanning homes, hospitals, factories, and vehicles. Consequently, the interconnected device count continues to surge at a rapid pace. These devices engage in communication with one another and with various services using diverse communication protocols to transport sensor or event data. These protocols facilitate applications in gathering, storing, processing, describing, and analyzing data to address a myriad of challenges. Additionally, IoT endeavors to furnish secure, bidirectional communication between interconnected devices like sensors, actuators, microcontrollers, or smart appliances, and corresponding cloud services. This study shows that into the evolution of Machine-to-Machine (M2M) protocol research, including MQTT, AMQP, and CoAP, spanning the past two decades, underscoring MQTT's pronounced growth trajectory. Furthermore, we compile pertinent application domains of MQTT, recognized as the predominant M2M/IoT protocol, via an exhaustive literature review across major digital research repositories. Here quantitative assessment illuminates key MQTT-related studies published in the last five years, facilitating a comparative discourse on MQTT protocol's principal attributes, advantages, and constraints. Additionally, we propose a taxonomy to juxtapose the properties and features of diverse MQTT implementations, encompassing brokers and libraries available in the public domain, aiding researchers and end-users in judiciously selecting a broker or client library tailored to their requisites. Ultimately, we finalize the pertinent outcomes of our comparison, spotlighting unresolved issues warranting further research.