Performance Modelling of Preemption-Based Packet Scheduling for Data Plane in Software Defined Networks

Software Defined Networking (SDN) is an emerging key innovative technology towards next-generation Internet. Through decoupling the control plane from the data forwarding plane, SDN is able to significantly simplify network management and accelerate flexible deployment. Many research efforts have recently been devoted to improving the performance of SDN. The majority of them, however, have focused on exploring the capacity of the logically centralised controllers in the control plane, but pay less attention to quantitatively investigating the performance of the underlying data plane. In order to fill this gap, this paper presents a preemption-based packet-scheduling scheme to improve the global fairness and reduce the packet loss rate in SDN data plane. Furthermore, an analytical model is developed to quantitatively evaluate this scheduling scheme and pinpoint the performance bottleneck in the SDN architecture. Both theoretical analysis and simulation experiments are conducted to demonstrate that this preemption-based scheduling scheme can achieve the better system performance compared to the traditional data plane packet scheduling in terms of global fairness index and packet loss probability.