A QoS-guaranteed intelligent routing mechanism in software-defined networks

Abstract

With the development of the Internet of Things (IoT), the network is required to guarantee the differential Quality of Service (QoS) requirements of the various data flows of various IoT services. Software-defined network (SDN) is envisioned as a promising technique to guarantee the QoS requirements of different services, through separating the control logic from data planes of networks. In order to guarantee the QoS requirements of data flows in SDNs, in this paper we investigate the problem of intelligent routing in SDNs, by leveraging a novel data flow classification method. Combining a variety of machine learning algorithms, a data flow classification method called MACCA2-RF&RF is proposed, in order to identify the data flow category and obtain the QoS requirements. The link parameter is newly designed considering multiple QoS requirements. According to the link parameter, the QoS-guaranteed path selection algorithm is then proposed, which can select QoS guaranteed routing path for different data flows with different QoS requirements. Aiming at the situation that the link is congested, local routing change algorithm is then proposed which only adjusts the links before and after the congested link instead of the entire path. Based on the above, a QoS-guaranteed intelligent routing mechanism called QI-RM in SDN is finally proposed in this paper to achieve QoS guarantee for data flows. The simulation results show that the MACCA2-RF&RF can classify data flows efficiently with an identification accuracy of 99.73%, and the QI-RM can guarantee the QoS requirements of data flows before and after link congestion.