The most popular messaging protocol in Internet of Things (IoT) networks is the Message Queuing Telemetry Transport (MQTT). It provides three Quality of Service (QoS) levels for reliable data delivery depending on application needs. Therefore, MQTT is widely adopted and supported by major IoT platforms like AWS IoT, Azure IoT Hub, IBM Watson IoT, and Google Cloud IoT. QoS enhancement techniques are needed in IoT networks because IoT devices and their applications operate in highly constrained and dynamic environments where standard communication protocols (like MQTT) alone cannot always guarantee reliability, timeliness, or efficiency. This study presents a comprehensive review of the recent and current literature on QoS enhancement techniques of MQTT-based IoT systems. These techniques are classified into five primary categories, which are; QoS enhancements at the protocol level, QoS enhancements at the network level, QoS enhancements at the broker level, cooperative subscriber approaches, and adaptive approaches based on context or machine learning. The related works belonging to these categories are explored and discussed in detail, focusing on the trade-offs of delivery reliability, protocol overhead, and energy efficiency. In addition, the study identifies the existing research gaps and limitations, and proposes future directions for advancing scalable, intelligent, and context-aware QoS management in MQTT-based IoT applications.
Nayef et al. (Sun,) studied this question.