DLICA: Deep Learning based novel Strategy for Intelligent Channel Adaption in Wireless SDN-IoT Environment

Software-Defined Network (SDN) is an emerging networking paradigm and provides the cost-effective scalability and flexibility required for an IoT environment. Here, wireless SDN is supposed to efficiently use limited network resources as well as provide smart traffic routing in order to deliver a bunch of IoT-generated data to the remote destination (Cloud). As, with the increasing number of IoT nodes and a huge amount of data generation, there is an immense load on SDN-IoT channels. This paper will mainly focus on appropriate channels assignment to each SDN-IoT enabled switch in order to prevent network congestion and improve network performance. Traditional fixed channel assignment algorithms in the SDN-IoT environments have failed because of large-scale and unpredictable dynamic traffic flows. In this paper, we have proposed a novel deep learning-based model called DLICA for intelligent channel assignment in wireless SDN-IoT. Moreover, a deep learning-based future traffic load prediction model has also been proposed to adaptively assign channels according to the traffic load prediction. Simulations show that proposed deep learning models achieve excellent accuracy while predicting traffic load and appropriate channel assignment. Apart from high accuracy, the proposed models also improve network performance significantly as compared to traditional channel assignment algorithms.