---

Financial transactions in Bitcoin are stored in a distributed database called the block chain. All transactions are publicly available for all network nodes with the aim of transparency and the possibility of verifying the correctness. But this blockchain transparency feature, exploited by transaction analysis techniques, can lead to the violation of users’ privacy and the disclosure of their identities. Researchers have proposed various techniques such as transaction mixing or fair exchange with the aim of improving privacy in Bitcoin transactions. In this paper, we present a new mixing scheme that overcomes some of the weaknesses of previous schemes. Obviously, in the proposed scheme, users can mix different amounts of Bitcoin in each round of the protocol implementation, which leads to achieving the result in a shorter time and at a lower cost. Also, this scheme is more resistant to denial of service attacks by malicious users.