End-to-End Delay Minimization Approaches Using Software-Defined Networking

The design and development of an efficient wide-area communication and computing approach remains as one of the greatest challenges to harvest the gigantic volume of data for a variety of application service needs in smart cities. The data may be of a varied priority due to the different stages of application services. This paper explores and leverages Software Defined Networking (SDN) to develop a communication approach for data transfers in smart city applications. Specifically, we propose an efficient approach to minimizing the end-to-end delay for data transmission in an SDN infrastructure by using a Timestamp Recording method to compare the arrival and departure of flows and packets over a period of time. The proposed SDN-based approach is designed to be QoS-aware. Hence, network traffic can be delivered according to the priority level of the traffic service. Finally, we evaluate the proposed SDN-based approach on the Global Environment for Network Innovations (GENI) testbed. We compare the Timestamp Recording method to other common delay measurement techniques, and our analysis demonstrates the effectiveness and the proposed SDN-based approach at scale.