Towards more reliable payment path finding via probabilistic modeling the uncertainty of channel balance

Summary:

The author of the letter shares research progress on achieving better payment path finding and reliability of the payment process in Lightning Network. The breakthrough is a simple model that puts the uncertainty of channel balances at its heart, which can be used to quantify the uncertainty in the channels and compute the expected number of attempts that are necessary to deliver a payment of a particular amount from one node to another participant of the network. Smaller amounts have higher success probabilities, and the success probability declines exponentially with the number of uncertain channels in a (multi)path. Sorting paths by their descending success probability during the trial and error payment process and updating the probabilities from current failures decreases the number of average attempts and produces a much faster delivery of payments. Additionally, if BOLT14 is implemented or nodes otherwise would pro-actively rebalance their channels according to previous research, the observed prior distribution changes from uniform to normal, and probabilistic path finding on a rebalanced network works even better. Implementations following the recommendation to use a probabilistic approach will tend to route payments along high capacity channels, making it harder to provide more liquidity. Nodes that provide a lot of liquidity and thus utility might be able to charge higher fees, allowing the emergence of a real routing fee market.Finally, the author believes to have a definite solution to the question of how to split a payment into k parts and how many funds to allocate to each path to increase the (multi)path success probability. He explains that an equal sized split as used in some implementations is not preferable, and there is way more math theory on how to actually solve the optimization problem in the general case and find a split and paths that maximize the probability of the attempts.

Updated on: 2023-06-03T03:52:44.306984+00:00