BDTTDM: A BLOCKCHAIN DUST TRANSACTION THREAT DETECTION MODEL USING HISTORICAL BIG DATA ANALYSIS AS A VACCINE

摘要

The existing approach to threat detection of dust transactions in a blockchain is conducted by setting a fixed amount of transactions. This is difficult to achieve due to the dynamic and variable nature of practical applications, so the performance is poor. This paper proposes a blockchain dust transaction threat detection model (BDTTDM) that uses the historical transactions generated in the blockchain as a vaccine for dust transaction detection. An artificial immune algorithm is used to detect dust transactions. The experimental results show that the dust transaction recognition rate by BDTTDM is 98.9%, which is 229.3% higher than that of the classic blockchain dust transaction model - Bitcoin Core client.