Simon Wadsworth's blog
Finally doing these again.
Got an implementation of PMIPv6 working in a virtualized environment.I was having trouble with tunnelling between two nodes, but this turned out to be an issue with kernel and then the Linux routing tables. Recompiling the kernel fixed most of the issues. I was able to validate my understanding of the protocol and how it functions by using a virtual switch (thanks to Brad for helping me with this) to configure vlans to make the mobile device appear to be connected constantly even though it was moving between access nodes. Essentially the mobile devices moves between access points or cell towers and then the nodes inside the core network perform Mobile IPv6 on their behalf. Was able to determine that even with tunnelling, I could still access parts of the network I shouldn't be able to. This was fixed with a simple firewall rule on the LMA node.
This week (starting Monday 12th) will a short week in the lab for me, so I'll need to do extra at home.
Answered some of the questions I had regarding the mobile core, which nows means I can attempt to define where I want to go a bit more. Might need to look at getting some hardware setup to actually look for potential attack vectors. My current thinking is that a firewall and correctly configured network shouldn't allow any traffic into secure areas, which would make core security a trivial exercise in network security. However this doesn't mean that there are no other attack vectors, which I will also need to investigate
As this week was a short week, I wasn't overly productive. Spent some time trying to answer some questions that I have regarding the protocols and network layout of LTE. I have only made some progress on this. Will need to do some more reading.
Spent some time setting up and configuring UML machines and installing and configuring them to act as an SGSN and GGSN in a GPRS network. Also had to configure a DHCP and DNS server on a third UML machine. This isn't specifically relevant to LTE as LTE does not use the same SGSN and GGSNs as in older GPRS networks, but it does allow me to use the GPRS Tunnelling Protocol (GTP) between them which may be of some use.
I also spent some more time reading and drawing diagrams to convince myself I understand the layout of the evolved packet system architecture, started to write summaries of the papers to make it easier to locate things in the future.
I spent the week rereading up on GPRS in GSM and UMTS networks and the Evolved Packet Core in LTE. I also had a look at some open source implementations of parts of a GPRS network to get an idea of how they work, and am hoping to get these running in a simple configuration next week