General question on routing difficulties



Summary:

PhD students at Purdue and the University of Waterloo have developed a novel routing algorithm called SpeedyMurmurs suitable for decentralized payment networks such as lightning network (LN). The approach uses an embedding-based methodology, where algorithms assign meaningful coordinates to nodes that enable efficient and effective discovery of payment paths.A path is then calculated in a flexible manner, with each intermediate node choosing the next node in the path as a function of its neighbors' coordinates, available funds, and closeness to the receiver. They have simulated several configurations of SpeedyMurmurs using real data from the Ripple network and compared it with other routing algorithms available in the literature. Their simulation results show that SpeedyMurmurs is able to find paths about twice faster, reduces the communication overhead by at least a factor of 2, and maintains a similar or higher payment success ratio. Privacy is also considered, and SpeedyMurmurs achieves value privacy, as well as sender and receiver privacy.In a discussion about the Lightning Network, Benjamin Mord suggested keeping an inventory of cryptographic primitives and unproven mathematical assumptions, as well as a resource to help the community respond quickly to cryptanalytic surprises. Christian Decker responded that routing is still in flux, but the network can easily scale to one million channels even on limited devices, and that upgrading to another protocol in the future is trivial. He also mentioned a recently proposed “revive” protocol for off-chain channel rebalancing. The Lightning Network is a first step towards reducing the load on the on-chain network, allowing timely on-chain settlements. Rebalancing is possible and can be achieved by disincentivizing the use of channels until they have been rebalanced.While there have been predictions that a hub-and-spoke network will form, it is extremely costly to create such a hub which will have to allocate sufficient funds to cover the maximum imbalance of all its channels ahead of time. The high fees associated with hub channels make them a bad choice for payments and nodes are likely to open bypasses that grab some of the traffic and associated fees from the expensive hub.However, lightning network centralization is not an existential risk as it is not one that affects the safety of the money supply itself. The lightning network seems extremely valuable and scalable regardless of off-chain payment rebalancing. Ultimately, the automations being built should encourage participants to open channels that support the network as a whole, not just themselves.


Updated on: 2023-05-24T03:15:43.733007+00:00