Abstract

Crypto currencies usage increasing every year around the world. The Bitcoin is the one of the famous cryptocurrencies, which is an unofficial usable currency in various nations. The bitcoin transactions are increasing, which needs to be monitored carefully. However, the conventional methods are failed to analyze the bitcoin transaction effectively. Therefore, this work focused on development of bitcoin transaction network (BTN) using pattern matching rules (PMR). Initially, the dataset preprocessing is carried out to identify the missed symbols, unknown characters from forensic blockchain dataset. Then, Petri-Net model applied on preprocessed dataset, which identifies the time stamp, transaction id, work tera hash, and work error properties. The Petri-Net model mainly used to parse and build the BTN model. Then, PMR conditions are developed to extract the transaction addresses extracted with time stamp details. So, PMR detects the illegal payment addresses by matching the known data with illegal (spam) addresses. Further, cache based PMR (CPMR) is also applied to detect the fraud transaction, which store all previous detected illegal payment addresses. So, for every new transaction, CPMR will ignore all those previously stored (detected) illegal payment addresses. This phenomenon causes reduction of fraud transaction detection time and processing becomes faster. The simulations shows that the proposed method resulted in reduced transaction processing time (TPT), fraud transaction detection time (FTDT), and improved fault transaction detection accuracy (FTDA) as compared to conventional methods.

Description