QoS-based Traffic Engineering in Software Defined Networking

Abstract

Quality of Service (QoS) is the overall performance of a computer network, particularly the performance seen by the users of the network. By managing the delay, bandwidth, and packet loss parameters, it allows us to allocate our available resources between applications in a reasonable way. In today's life, it is general that many applications run at the same time. As a critical type of resource, bandwidth would be shared without principle which leads to interference. Therefore, “important” applications cannot get enough bandwidth to transmit data. To solve these problems, the end-to-end bandwidth resource allocation scheme is proposed to support Quality of Service (QoS) for various types of traffic based on the user QoS demand. The main goal of this system is to provide high QoS performance for high priority flows. In order to meet this goal, this system considers the flow priority and the dynamic characteristics of the network. In addition, feasible paths are calculated for all the traffic flows that can satisfy the user bandwidth demands. In order to mitigate the flow performance degradation and congestion, the controller checks the link bandwidths by reserving the required bandwidths for incoming flows. If the link bandwidth is smaller than the predefined threshold value, the link is defined as the bottleneck link and reroute the highest priority flow from the bottleneck link to an alternative link that has enough bandwidth for the rerouting flow. Furthermore, to improve the performance and to ensure the QoS of the high priority flows, a queue mechanism is used which is provided by the OpenFlow at the data link level. This research will try to accommodate as much traffic as possible, and study the effect of routing on a rather general mix of QoS traffic types. The effectiveness of the proposed scheme is described on the emulated SDN network. The proposed scheme is compared with the conventional shortest path scheme, multipath routing scheme in the various network topology. Furthermore, the performance of the proposed scheme is compared with the popular flow scheduling scheme Hedera in the data center network topology. The improvement of QoS traffic type is quantified in terms of throughput, delay, jitter and packet loss rate, respectively. Based on the experiments, the researcher observed that the proposed method (QoS_based Traffic Engineering Approach) QT offers a significant improvement compared to a static, traditional IP network and the multipath network environment byiv providing better performance in terms of packet loss rate for the QoS traffics and great improvement in link utilization.