Abstract:
This paper explores a method for improving the data transmission process in resourceconstrained IoT systems that rely on protocols such as MQTT, CoAP, and LwM2M. Drawing on recent analyses of the strengths and weaknesses of these widely used application-layer protocols, the study identifies a key research gap: very few papers concentrate on different application layer protocol interconnection, and no research is present on IoT transmitters capable of dynamically selecting which protocol to use under changing network conditions. To address this need, we propose a system design that supports devices switching between two different application-layer protocols without incurring the substantial overhead of full firmware updates. Instead, devices receive updated configuration profiles over the network, thus minimizing downtime and resource consumption. A dedicated backend infrastructure coordinates the protocol translation and gateway functions and collects network indicators (e.g., signal-to-noise ratio and retransmission counts). These indicators feed a decision-making module, which adjusts the configuration parameters sent to IoT transmitters. This arrangement is expected to reduce energy usage and improve reliability for battery-powered, low-power devices operating in unstable environments by using the most suitable application protocol for transmission at a given moment. Future developments will incorporate advanced analytics, including machine learning, to refine the decision process and further enhance adaptability and scalability.