Improving Machine to Machine Communication in Integrated MQTT and COAP IoT Network

Abstract

This study presents a pioneering integration of MQTT and CoAP protocols within a Software-Defined Networking (SDN) framework, complemented by a Support Vector Machine (SVM)-based flag status indicator, to enhance interoperability and performance in IoT networks. The primary goal is to address the escalating demand for seamless communication in heterogeneous IoT environments. The integration of MQTT and CoAP through the SDN controller signifies a significant advancement, providing centralized control and programmability for dynamic management of communication protocols and network resources. This approach fosters flexibility, essential for accommodating diverse IoT devices with varying communication requirements, streamlining communication pathways and elevating network efficiency. A key innovation is the introduction of an SVM-based flag status indicator, leveraging SVM's capabilities in handling non-linear relationships and resisting over fitting. This indicator proves instrumental in accurately determining flag status, surpassing alternative classifiers like Decision Trees, Naive Bayes, and Random Forests. The study's experimental results validate the success of the integrated approach, demonstrating superior throughput, sustainable residual energy levels, reduced end-to-end delays, and a higher message delivery rate compared to alternative classifiers. These outcomes affirm the methodology's efficacy in achieving seamless interoperability, efficient communication, and robust decision-making across diverse IoT scenarios.