The Bitcoin network is a relatively adolescent network, as it was created in 2009. It has continued to grow and garner more attention ever since. Certain events in Bitcoin’s history, like the opening of public trading with USD and the announcement of SatoshiDice have led to a noticable increase in usage.

The purpose of the Bitcoin network, however, remains constant: it is an anonymous peer-to-peer network. The users value their own privacy and the privacy of their transactions, which creates challenges for researchers aiming to study the growth of the network. As such, it is difficult to abstract meaningful data, as the network has largely been successful in its goal. However, watching the graph evolve presents a unique opportunity: we can observe how traits in the graph change over time and how this reflects the
overall mission of the network.

In this paper, the authors model the Bitcoin transaction graph in small time slices using several different models, including one of their own. In order to gauge the effectiveness of each model, they compare some of the attributes of the generated graphs, like diameter, clustering, etc. While their model captures some features of the Bitcoin graph, none of the models fully capture its complexity. They also present a possible algorithm for detection of money laundering, and show that it does detect abnormalities present in Bitcoin but not present in their models.

Read the research paper to find out more!