Effective Routing in Payment Channel Networks using Spanning Trees

Abstract

In today’s world, the demand for cryptocurrencies is increasing, making scalability a significant challenge. Payment Channel Networks (PCNs) have emerged as a viable solution. However, PCNs still face challenges in routing payments effectively. Payments typically flow over a single channel in one direction, leading to channel depletion and failed trans- actions in that direction. Consequently, na ̈ıve routing schemes, such as shortest path routing, can exhaust critical channels and freeze the sys- tem. To address these issues, we present STR (Spanning Tree Routing), an algorithm that leverages the entire network and rebalances channel capacities dynamically. We are able to route large transactions over low- capacity channels while utilizing a larger portion of the network. Our experimental studies show that STR performs significantly better in net- work scenarios of low demand density, where traffic predominantly flows in one direction when compared with the shortest path algorithm. In such scenarios, STR algorithm provides performance improvement by a factor of 50 compared to the shortest path algorithm and also leads to fewer imbalanced channels.