Adversarial behaviours in mixing coins under incomplete information

Yilei Wang,Andrea Bracciali,Guoyu Yang,Tao Li,Xiaomei Yu

DOI: https://doi.org/10.1016/j.asoc.2020.106605

IF: 8.7

2020-11-01

Applied Soft Computing

Abstract:<p>Criminals can launder crypto-currencies through mixing coins, whose original purpose is preservation of privacy in the presence of traceability. Therefore, it's essential to elaborately design mixing polices to achieve both privacy and anti-money laundering. Existing work on mixing policies relies on the knowledge of a blacklist. However, these policies are paralyzed under the scenario where the blacklist is unknown or evolving. In this paper, we regard the above scenario as games under incomplete information where parties put down a deposit for the quality of coins, which is suitably managed by a smart contract in case of mixing bad coins. We extend the <em>poison</em> and <em>haircut</em> policies to incomplete information games, where the blacklist is updated after mixing. We prove the existence of equilibria for the improved polices, while it is known that there is no equilibria in the original <em>poison</em> and <em>haircut</em> policies, where blacklist is public known. Furthermore, we propose a seminal <em>suicide</em> policy: the one who mixes more bad coins will be punished by not having the deposit refunded. Thus, parties have no incentives to launder money by leveraging mixing coins. In effect, all three policies contrast money laundering while preserving privacy under incomplete information. Finally, we simulate and verify the validity of these policies.</p>

computer science, artificial intelligence, interdisciplinary applications