
The control plane instructs functionality whereas the data plane performs the functionality instructed by the control plane. In SDN, the control plane is separated from the forwarding devices. This has several benefits. For instance, management of forwarding devices becomes easier as there will be an overall image of the network at the remote controller. In addition, the internal architecture of forwarding devices becomes less complex as they need not store massive amount of state information. They will be controlled and programmed via a logically central controller through standardized interfaces. Research on SDN involves research on the control and data planes as well as the standardized interfaces.
In his dissertation, Hesam Zolfaghari deals with the hardware architecture aspects of the data plane. He looks for an answer on how to make the data plane more flexible and capable of supporting higher rates of packet arrival while keeping the hardware complexity low.
“What is meant by flexibility in this context is providing the means to adapt to new and upcoming networking protocols just by means of software upgrades rather than hardware upgrades”, Zolfaghari explains.
These goals at first seem contradictory. However, Zolfaghari’s research proves that they can be achieved.
“The key to this achievement is analysing the application domain effectively and making the right architectural choices. The results of the dissertation are not only of benefit to architects of packet processing hardware, but to network protocol designers as well”, Zolfaghari says.
The doctoral dissertation of M.Sc (tech) Hesam Zolfaghari in the field of Computing and Electrical Engineering titled Flexible Low-Area Hardware Architectures for Packet Processing in Software-Defined Networks will be publicly examined in the Faculty of Information Technology and Communication Sciences at Tampere University at 12 PM on Monday 21.12.2020 online. The Opponents will be professor Peeter Ellervee from Tallinn University of Technology and associate professor Guido Maier from Politecnico di Milano. The Custos will be professor Jari Nurmi from the faculty of Information Technology and Communication Sciences.
The dissertation is available online at the http://urn.fi/URN:ISBN:978-952-03-1806-2
