Präsentation Bachelor Thesis von Matthias Lobenhofer: „Anonymity Analysis of JoinMarket“
Am 22. Dezember 2022 um 17:45 Uhr präsentiert Matthias Lobenhofer seine Bachelor Thesis zum Thema „Anonymity Analysis of JoinMarket“.
Zur Teilnahme via Zoom verwenden Sie die folgenden Daten:
https://fau.zoom.us/j/69294317201?pwd=YW5DZDdMSlRIOGtoQzVQVllzRlFvdz09
Meeting ID: 692 9431 7201 Passcode: 455246
Unten stehend finden Sie die Kurzfassung des Vortrags in englischer Sprache.
Abstract:
Bitcoin is probably the most famous cryptocurrency today with hundreds of thousands daily transactions. It is often perceived as an anonymous payment method. However, it is a fallacy to think Bitcoin is anonymous, as it does not contain names or user accounts. The complete public transaction history in the Bitcoin blockchain contains many small bits of information about its users. With blockchain analysis tools like the common-input-ownership heuristic these small bits can be combined to deanonymize Bitcoin users. A lack of privacy might have undesirable consequences for them. To prevent the accumulation of their private information, Bitcoin users can employ services like JoinMarket. JoinMarket brings together privacy seeking people so they can obfuscate their tracks through joint transactions, which are called CoinJoins. The objective of this bachelor thesis is to analyze the anonymity provided by JoinMarket and to determine possible flaws.
In order to create a method to identify JoinMarket transactions in the blockchain we work out their specific characteristics. This is not a straight forward task as other transactions share a lot of similarities with JoinMarket transactions. We modify and extend an existing identification approach with multiple new requirements to achieve a better precision in detecting JoinMarket transactions. Our results show that JoinMarket is still widely used today, despite new providers of similar services. Furthermore, we detect a weakness in the timing between CoinJoins. Under certain assumptions, which we know are wrong in 14% of the cases, we are able to identify one of the participants with a 50% chance. Our results do not render JoinMarket unusable. JoinMarket CoinJoins generally remain not resolvable, but there are damaging implications for some of the involved. Users should be aware of how their actions influence their anonymity.