Enhancing Software-Defined Networks with Intelligent Controllers to Improve First Packet Processing Period

Citations Over TimeTop 15% of 2023 papers

Abstract

Software-Defined Networking (SDN) has a detailed central model that separates the data plane from the control plane. The SDN controller is in charge of monitoring network security and controlling data flow. OpenFlow-enabled routers and switches work as packet-forwarding devices in the network system. At first, OpenFlow forwarding devices like routers and switches do not know how to handle the data packets transmitted by the host. This is because they do not have any security controls, policies, or information. These packets are sent to their destination. In this situation, the OpenFlow forwarding device sends the first data packet of a host to the SDN controller, which checks the control packets for the data packet and creates flow entries in the switch flow table to act on the following categories of data packets coming from the host. These activities at the SDN controller and switch levels are time-intensive, and the first data packet from the host always takes a longer time to reach its destination. In this article, we suggest an SDN controller with instant flow entries (SDN-CIFE) to reduce the amount of time it takes for the host to transmit its first data packet. Before traffic comes from the host, our method adds the necessary flow entries to the flow table of the OpenFlow switch. The technique was made in Python and tested on a Mininet network emulator using the RYU controller. The results of the experiment show that the time it takes to process the first data packet is reduced by more than 83%.

Related Papers

• → A Survey on Data Plane Flexibility and Programmability in Software-Defined Networking(2019)87 cited
• → An OpenFlow Extension for the OMNeT++ INET Framework(2013)42 cited
• → An OpenFlow Extension for the OMNeT++ INET Framework(2013)24 cited
• → User-defined actions for SDN(2014)10 cited
• → A survey on issues of concern in Software Defined Networks(2015)8 cited