Metaheuristic deep neural network-based Intelligent Routing in SDN

The software-defined network is an architecture that controls and deals with the entire network utilizing the control, the foundation layer, and the application layer. Traffic prediction at present can be absorbed and estimated by the SDN (Software Defined Network) but the future traffic cannot be predicted by SDN. Many types of research have been done using LSTM (Long Short Term Memory) for predicting traffic. LSTM (Long Short Term Memory) model provides the futuristic accurate output predictions with the help of current information provided. m-LSTM provides improved results in recurrent function with maximum data flow in the performed task. The computational technique of swarm intelligence is performed for solving complex problems by collecting the data on individual performances and behavior. The problem is traffic prediction in the dynamic routing for SDN. This paper proposes the significance of predicting the optimal route and traffic using DC m-LSTM in the recurrent neural network which is the novel approach for SDN to predict the future traffic in the optimal route. The best route is identified based on the traffic prediction using the metaheuristic DC m-LSTM swarm intelligence.