A Balanced Routing Algorithm for Blockchain Offline Channels using Flocking
0 likes449 views
This document proposes a flocking-based routing algorithm for offline blockchain channels to balance load. It converts channels to nodes in a network and assigns coordinates. Channels move away from overused channels (repulsion zone) and toward underused channels (attraction zone). The coordinate updating process aims to evenly distribute channel usage. Evaluation on Bitcoin Lightning Network data shows the algorithm maintains low standard deviation of channel balances compared to shortest path routing and completes more payments with shorter paths on average.
![Angle of the channel
Coordinate of each channel
is in positive numbers
Hence, the angle of each
channel is in the range
[0,90]
25](https://image.slidesharecdn.com/blockchain2019copy-190626195931/85/A-Balanced-Routing-Algorithm-for-Blockchain-Offline-Channels-using-Flocking-25-320.jpg)
A Balanced Routing Algorithm for Blockchain Offline Channels using Flocking
- 1. Subhasis Thakur and John Breslin Data Science Institute National University of Ireland Galway A balanced routing algorithm for blockchain offline channels using flocking
- 2. - - SFI Research Centre - Precision pasture-based dairying (agtech) - Blockchain applications in the dairy supply chain 2
- 5. 5
- 6. 6
- 7. 7
- 8. 8
- 9. 9
- 10. 10
- 11. 11
- 12. 12
- 13. 13
- 14. 14
- 15. 15
- 16. 16
- 17. 17
- 18. 18
- 19. 19 Load balancing problem in the Lightning Network
- 20. 20 Privacy-preserved load balancing problem How can we ensure uniform usage of channels without revealing channel balances?
- 22. Step 1: Convert channels to nodes of a network Step 2: Assign coordinates to each channel 22 (Vj ,Vb )
- 23. Flocking of channels 1. A set of channels moves away from each other if used in a PBT 2. The set of mirrored channels moves close to each other 23
- 24. 24 Flocking of channels Zone of Repulsion Zone of Alignment Zone of Attraction
- 25. Angle of the channel Coordinate of each channel is in positive numbers Hence, the angle of each channel is in the range [0,90] 25
- 26. Coordinate updating 26 Angle1 = 180 degrees + mean angle of the channels in the repulsion zone (opposite of channels in the repulsion zone) Angle2 = Mean angle of the channels in the alignment zone Xi = Xi + 0.2(Cos(Angle1 ) + Cos(Angle2 )) Yi = Yi + 0.2(Sin(Angle1 ) + Sin(Angle2 ))
- 27. Routing protocol 27 ● Updated coordinates are regularly announced ● Routing as per the shortest path
- 28. 28 Bitcoin Lightning Network Data used = three subgraphs of BLN # of nodes Average degree Average channel balance (simu- lated) Average PBT value (simu- lated) 400, 600, 800 11 7 0.5
- 29. Network with 400 nodes The standard deviation of channels remain low for balanced routing 29
- 30. Network with 600 nodes The standard deviation of channels remain low for balanced routing 30
- 31. Network with 800 nodes The standard deviation of channels remain low for balanced routing 31
- 32. The number of completed PBTs for balanced routing is higher than the same for shortest path routing Average path length for PBTs is lower for balanced routing 32