ckuemmerle
Christian Kümmerle

https://webpages.charlotte.edu/ckuemme1/

Data-efficient machine learning, optimization, Bitcoin, Lightning network. Assistant professor in Computer Science at UNC Charlotte.

Public Key

npub1amv8kzuluqq50rxny6klwslm6rpkdyusedcse6wt2dh0m5zqresqd3w3v4

Profile Code

nprofile1qqswakrmpw07qq283nfjdt0hg0aapsmxjwgvkugva894xmha6pqpucqpzemhxue69uhhyetvv9ujumn0wd68ytnzv9hxgspyyn8

Publishing to

relay.nostr.band

Author Public Key

npub1amv8kzuluqq50rxny6klwslm6rpkdyusedcse6wt2dh0m5zqresqd3w3v4

Show more details

Last Notes

2024-10-26 13:28:40

Today between 10:00-10:30 am at @npub17yq…5269 , learn about our new research on Lightning Network pathfinding strategies from my great student @npub1yav…dqdm Saraswathi https://github.com/orgs/TABConf/projects/4?pane=issue&itemId=77244631&issue=TABConf%7C6.tabconf.com%7C109
- We present a detailed breakdown of the underlying cost functions and constraints used by pathfinding algorithms of the different LN node implementations.
- Surprisingly, we observe that the pathfinding problems that most LN node implementations attempt to solve are NP-complete, and cannot be guaranteed to be optimally solved by the variants of Dijkstra’s algorithm currently deployed in production.
- Through comparative analysis and simulations, we evaluate efficacy of different pathfinding strategies across metrics such as success rate, fees, path length, and timelock.
We think that this study provides new insights for how pathfinding within the Lightning Network can further be improved. While beyond the scope of our study, the findings are also relevant for multi-part payments.
Link to paper: https://arxiv.org/pdf/2410.13784