Adaptive Software-defined Network Controller for Multipath Routing based on Reduction of Time

Abstract

Software-defined network (SDN) is a new paradigm in the networking that makes a programmability and intelligence the networks. The main SDN characterize is separating network management (control plane) from the forwarding device (data plane). SDN logically centralizes the network with the programmable controller which collects global knowledge about the network. The SDNs can improve the performance of the routing packets in the networks because of agility and the ability to create a policy for a driven network. In the multipath routing, the SDNs controller is responsible to calculate the routes of optimum path and alternative path wherever a link is failed. However, a high delay time calculation of selecting optimum and alternative paths in multipath routing by the SDN controller is observed in the recent investigations. In this paper, we propose an efficient algorithm for SDN multipath routing controller. The mechanism of the proposed approach calculates the best path from the source to the destination which is based on using adaptive packet size and observing network link capacity. The proposed algorithm considers reducing delay time of the link handling when the flow traffic switches from the main path to the recovery path. As a result, this approach is compared to some state of the arts according to the delay time of choosing the best path and alternative paths in a given network topology. SDN based on the proposed algorithm consumed approximately 1 ms for selecting recovery routes. On the other hand, the proposed algorithm can be integrated to an SDN controller which provides better consolidation of transmission for sensitive applications as video streaming.