Adaptive Channel Assignment Method For SDN Enabled IoT

Abstract

With the advent of Internet of Things (IoT), the transmission speed has become a vital issue to be
addressed. Assigning appropriate channels to the traffic in SDN enabled IoT networks can impact
transmission speed tremendously. However, there exists many conventional methods, the unpredicted
traffic load make these conventional methods unsuitable to accommodate the dynamicity of IoT
environment. Recently, the Software Defined Networking has been emerged as supporting
technology to enhance the performance of IoT networks and to improve the quality of transmission.
Still, there is a further scope for dynamic and machine learning techniques to predict the load of
network traffic and to allocate the channel for better assignment of resources and better utilization
of under-utilized channels. Hence, we are proposing deep neural networks based traffic load
prediction and channel allocation techniques for improving throughput of the channels. The
proposed methods are smart enough to accommodate the dynamic traffic load by predicting it in
advance and quickly assign the appropriate channels in SDN enabled IoT for better resource
allocation. The performance of the algorithms is also contrasted using statistical performance
matrices