In the past few years SDN (Software Defined Networking) and OpenFlow became extremely popular.
The goal of the OpenFlow standard is to allow researchers to try experimental packet forwarding mechanisms in real networks. To make it easily adoptable, the standard was designed to make use of existing hardware and let networking equipment vendors implement it without exposing the internal working of their switches.
Because of this some compromises had to be made. The OpenFlow switches have limited capabilities, which makes implementing unorthodox routing protocols on them hard.
The aim of my thesis is to examine the possibilities of extending the action set of OpenFlow with new hardware accelerated operations which can support these new packet forwarding methods.
In my paper I also introduce the problematic Bloom filter based and Greedy (geographic) routing algorithms. I write about NetFPGA (a developer platform designed to help implementing network algorithms running at line rate) and examine how it can be used to extend the capabilities of OpenFlow networks. I implement the mentioned algorithms and test them to see if they are suitable to use in real world scenarios.