Network Slicing in 5G Connected Data Network for Smart Grid Communications Using Programmable Data Plane

Thumbnail Image
Journal Title
Journal ISSN
Volume Title
Sähkötekniikan korkeakoulu | Master's thesis
Communications Engineering
Degree programme
CCIS - Master’s Programme in Computer, Communication and Information Sciences (TS2013)
Due to the technological advancements in communications, contemporary smartgrids have started to adopt Fifth Generation (5G) mobile networks for communications. Communication between Supervisory Control and Data Acquisition (SCADA) systems and Remote Terminal Units (RTUs) in smart grid environment utilizes the IEC 60870-5-104 protocol. It is a Transmission Control Protocol/Internet Protocol (TCP/IP) based protocol where data is transmitted in unencrypted form. Smart grids adopting 5G networking for communications are not isolated appropriately. Therefore, smart grids are still insecure against cyberattacks. With respect to recent developments in data plane programming, new networking paradigms can be realized including progressive ways of isolating smart grid traffic from normal traffic in a data plane. The aim of the thesis is to explore the usage of data plane programming to isolate and secure smart grid traffic into a network slice in 5G networks. This thesis successfully develops a flexible and efficient 5G network slicing solution based on P4 (Programming Protocol-Independent Packet Processors) language framework. Slice isolation is achieved with varied packet rates in slices as well as blocking devices from one slice communicating to the devices in another slice. The network slicing solution enables 5G equipped RTUs to be connected with SCADA in the Data Network in an isolated manner. A P4-based packet tagging solution is also presented where smart grid packets are tagged with specific Differentiated Services Code Point (DSCP) in the Internet Protocol (IP) headers to aid network slicing. DSCP values in the IP headers are used by the P4-based slicing solution to classify smart grid packets appropriately and push them into network slices. Both the network slicing and the DSCP tagging solutions are implemented with P4 software switch known as the Behavioral Model version 2 (BMv2). The network slicing performance is assessed in an experimental 5G testbed, which is powered by an opensourced 5G core. Basestation and User Equipment (UE) elements for connecting the RTU are simulated using appropriate software. The network slices are examined carefully in this thesis as well as their ability to provide Quality of Service (QoS) for the services hosted in the slices.
Kantola, Raimo
Thesis advisor
Ghasemi, Mandana
P4, SDN, network slicing, 5G, smart grid, data plane programming
Other note