Method of Bitcoin transaction output consolidation using spending simulation

Abstract

In the Bitcoin protocol, each transaction consumes transaction outputs – records that associate a specific amount of bitcoin with a particular owner – via transaction inputs, which are references to existing outputs, and creates new outputs with new amounts and owners, thus reflecting an act of ownership transfer. Users who frequently receive payments accumulate a large number of small unspent outputs, which increases costs of future transactions, decreases performance of wallet software, and threatens financial privacy. In order to reduce the number of outputs, users perform consolidation operations that merge many small outputs into a larger one. However, such transactions or their cascades are easily identifiable in the transaction graph, which allows linking the outputs involved in a single consolidation operation and thus deteriorates the user's financial privacy. This paper proposes a method of multi-step output consolidation through simulated natural spending. It uses simulating transactions with multiple inputs and two outputs which resemble typical payments with a target output and a change output. The proposed method iteratively reduces the set of user’s unspent outputs to a certain target size without creating obvious patterns in the transaction graph. Cost analysis demonstrates that even when the user sacrifices efficiency in favor of minimizing the size of linked output groups, the total cost of multi-step consolidation increases only by a factor of 2-4 compared to simple consolidation. The article also outlines future research directions, including the identification of transaction graph patterns that may indicate the use of complex multi-step consolidation schemes, as well as patterns of reverse operations of splitting outputs into smaller parts, which may indicate illegal activities