Design and Realization of Cognitive Routing Resources Using Big Data Analysis in SDN

Future networks need to get knowledge of users´ requirements by analyzing big data to achieve intelligence. OpenFlow is the first standard communications interface defined between the control and forwarding layers of a SDN architecture. So as to enhance network flexibility and availability, this paper presents a network resource allocation scheme equipped with the ability of user preference awareness, by connecting a cloud platform of analyzing users´ data. We use the Hadoop platform to predict the rate and type of the incoming flows and employ them to get the lightest burden link through the LARAC algorithm. It has been verified on the BUPT SDN test bed that this system is capable of allocating the network resources dynamically by predicting the network load in advance. The results show that the proposed methods achieve better load balance than previous networks.