Anti-Dust: A Method for Identifying and Preventing Blockchain’s Dust Attacks

As a global decentralized distributed ledger, Bitcoin’s POW mechanism is so complex that the transaction’s confirmation is very time-consuming. If the dust attacks occur in the blockchain, which sends large amounts of dust transactions to the blockchain, the time of transactions’ confirmation will be more longer, which will lead to more security problems, such as double-spending or blockchain forks. In order to identify and prevent dust attacks, this paper proposes an innovative Anti-Dust security model, which analyzes the characteristics of dust attacks and learns from the historical transactions big data, uses the Gaussian distribution to construct a dust recognition algorithm. At the same time, a dust transaction pool will be added to improve the blockchain architecture. In the experiment, a large amount of dust transactions will be send to the blockchain. Taking 50000 dust attacks and 500 normal transactions as an example, the conformation time of blockchain without Anti-Dust is about 20000 seconds, and Anti-Dust is about 210 seconds. The results show that Anti-Dust is an effective defense method to identify and prevent the dust attacks.