Intelligent Routing Approach based on Machine Learning and SDN for Heterogeneous IoTs

Abstract

Internet of things (IoT) is a network of billions of objects connected through heterogeneous communication technologies such as Ethernet, WiFi, Bluetooth, Zig-Bee, etc. The diversity of communication technologies is a significant barrier preventing the management, the control and more specifically the interoperability using a unified interface. In this paper, the SDN layer, which aims at ensuring a universal management of the IoT heterogeneous network, is to develop a Machine Learning algorithm that allows the classification of IoTs according to the data, this algorithm allows the COBOX to find the object that will be interested in the data sent by other IoT, finally integrate this algorithm in the SDN network. The proposed solution enables the building of a middleware that supports interoperability between heterogeneous devices.