A year ago I set out to test various existing mesh routing protocols on different topologies. Especially emulating huge networks are tempting, since real hardware for this is too expensive (for me). I present the story, the projects [1] and my findings while testing emulated mesh networks of up to 2000 nodes.

Protocol implementations tested:

• Babel
• batman-adv
• BMX6
• BMX7
• CJDNS
• OLSRd
• OLSR2
• OSPF

[1] https://github.com/mwarning/meshnet-lab

I like to show you what the problem of mesh routing ist and how you could build your own mesh routing protocol that is good enough.

Central Questions:

• What is Routing?
• What is a mesh network?
• Why is this an interesting problem?
• How to get from A to B without asking everybody?

This talk will go into more details of mesh routing.

• What are metric and who cares?
• What are the challenges when using WiFi?
• What are the limits of mesh routing?