Machine Learning and Zero Knowledge Empowered Trustworthy Bitcoin Mixing for Next-G Consumer Electronics Payment

Abstract

In the next generation of consumer electronics, bitcoin mixing scheme is an essential part to realize decentralized anonymous payment. However, there are still some challenges in existing decentralized schemes. First, existing schemes necessitate broadcast of sensitive information for group construction and lack of machine learning models to make assisted decisions. Second, these schemes fall short in verification of consumers’ integrity prior to their admission. Last but not least, the decentralized protocols tend to lack robust mechanisms for protecting the details of transactions during negotiations. To address the above challenges, this paper proposes a machine learning and zero knowledge empowered trustworthy bitcoin mixing for next-G consumer electronics payment to enhance consumer privacy and transaction details protection. Specifically, we design a mechanism based on zk-SNARKs for verifiable proofs to preserve privacy. Moreover, we construct a model based on machine learning to assist in decision making and verify the integrity by Pedersen commitments. Finally, proposed scheme refines an approach to guard transaction details during negotiations. The experiment demonstrates our approach offers enhanced efficiency and anonymity assurances without sacrificing performance.