Research Area: Software Defined Networks
Software Defined Networks (SDN) is a paradigm in which control and data planes of traditional networking devices are decoupled to form a distributed model. Communication between the separate planes requires a protocol such as Open Flow to leverage programmable routing and forwarding decisions on the network. In this model, Application Programmable Interfaces (APIs) make it possible to inject policy and forwarding rules via the control plane or controller. The most prominent challenges resulting from the separation is link security between the separated elements through which private network data is now traversing.
One main area of concern is the method of transmission with which the majority of Open-Source controllers currently communicate. The preferred practice is for a Transport Layer Security (TLS) channel initiation by an Open Flow switch wishing to communicate with a controller. Many developers have replaced the TLS method of communication with straight Transport Control Protocol (TCP) due to handshake sequence issues caused by certificate exchange during the TLS connection phase.
This thesis and the subsequent research will ask questions on security around the controller to device links that pass flow tables , network abstractions and multi-layer information to multiple controlled network elements.
The main objective of this research is to develop testing procedures that allow for accurate and repeatable experiments. Therefore, in researching security vulnerabilities between controllers and forwarding devices, bench marking performed on secure links tests the capability of authentication mechanisms to function properly under load.
The outcomes of this research include a series of quality industry standard tests to benchmark typical SDN controllers and forwarding devices. A critical analysis of typical devices at low, medium and high loads. An SDN security taxonomy is presented to help with future categorizing of device testing in context of SDN architecture.
Name of the Researcher: Peter Edge
Name of the Supervisor(s): Zhang, Zhongwei, Lai, David
Year of Completion: 2019
University: University of Southern Queensland
Thesis Link: Home Page Url
