BTC2019

In this rather research oriented session, we take a data science perspective and focus on the network part of the Lightning Network. After a short introduction of the Gossip protocol we discuss the issues with pathfinding and the incentives for nodes participate in routing. We will understand how the betweeness centrality score is roughly proportional to the expected earned routing fees of a node. We discuss how recommendation engines like the lib_autopilot.py can help us to create a robust network topology that increases the success of routing and decreases the difficulty of path finding. We will also look at other path finding strategies like JIT-Routing, Atomic Multipath Routing, trampoline payments as well as some ideas for congestion and flow control.

Since this session is research oriented the concrete topics are only tentative and might be changed on short notice and in response to ongoing trends.

This course is appropriate for current and aspiring developers who want to understand routing.

Prerequisite: none.