An analysis of on-chainLightning Network transactionsin the Bitcoin blockchain
MetadataShow full item record
The blockchain is one of the main mechanisms enabling Bitcoin to be a decentralized electronicpayment system. It provides the system with a shared transaction history, which can be verified byevery participant. With increased use, it has become apparent that there is limited possibility forscaling this to handle more transactions. Payment channel networks are one proposed solution tothis problem. They allow for more transactions to be done by moving some transactions to a sepa-rate network. The Lightning Network is one such network, which uses Bitcoin and the blockchain tooperate. While the transactions in this network will not be included in the blockchain, there will bedata there related to the network. This is because it needs the blockchain to manage the paymentchannels which the network consists of.In this project we have explored the blockchain with the goal of identifying transactions relatedto the Lightning Network, and by doing so, determine what information about it is available in theblockchain. We have created different methods for identifying these transactions. The methods usedifferent transaction characteristics differing in uniqueness, making some methods more precise,but having fewer results, and vice versa. We created software implementing the methods, whichwere used to parse the blockchain. The effectiveness of these methods have been quantified bycomparing the data we found when parsing the blockchain, to data we collected directly fromthe Lightning Network. The results shows that the methods are viable for identifying a subset oftransactions, and that precision can be sacrificed for finding more. By identifying these transactionswe were able to determine what information about the Lightning Network we can see from theblockchain perspective, and also some aspects where we are limited.We have also adapted heuristics from previous work doing blockchain analysis to our scenario.These were used to link related information we had found when parsing the blockchain, whichenabled us to create network graphs showing the relations between the Lightning Network channelsidentified on the blockchain. While the relations in this network graph were limited, comparedto the actual relations found within the lightning network, they show how the blockchain canbe used to infer non-explicit information about the lightning network. We have also identifiedseveral methods for potentially inferring or locating more information using what is available inthe blockchain.