Secure and Efficient Decentralized Payment Systems Using Blockchain and Advanced Fraud Detection

Abstract

The problem of fraud transactions detection in digital payment systems is relevant due to the wide spread of fraud transactions as a consequence of increased traffic and threatens the financial security and trust. The traditional fraud detection methods are also prone to problems of scalability and cannot scale well with new emerging cyber threats. Based on blockchain technology, this research was done and suggested secure and efficient decentralized payment system (DPS). It is integrated with Isolation Forest algorithm, this is a machine learning based anomaly detection technique which comprises the system for isolating the fraudulant transactions in real time.

The blockchain based DPS architecture proposed has a feature of immutability and decentralization to avoid conducting any type of fraud which can be done in the financial system. It is analyzed through a complete experimentation along with a real worldly transaction dataset with providing high accuracy in fraud detection at a reasonable transaction speed and scalability. It is equally clear as compared to the existing fraud detection models in order to avoid false positives as much as possible by low false positive rate, and as much as possible in the new fraud pattern learned by low training query rate.

Study present a sound framework for fraud detection that allows us to tackle core security as well as efficiency issues in DeFi and meet existing security and regulation requirements. In these results, financial institutions, policymakers, as well as researchers gain insight into the problems posed regarding privacy, scalability, and security in building decentralized payment infrastructures