Blockchain oracle networks serve as critical bridges between on-chain smart contracts and off-chain data sources, enabling decentralized applications to access real-world information. However, existing oracle systems suffer from significant vulnerabilities including data manipulation attacks, lack of quality assurance mechanisms, and absence of robust validation frameworks. This research proposes a novel reputation-based data quality assurance system for blockchain oracle networks that combines machine learning-based reputation scoring with stake-weighted validation mechanisms. We employ a multi-tier validation process to check data sources for historical accuracy, metric consistency and behavioural patterns. Through extensive simulated and experimental studies with 1,000 oracle nodes under different data categories, we show a reduction of 82.3% in false data injection attack exposure as well as an improvement by 76.8% in our overall data quality metrics compared to conventional oracle systems. The proposed system achieves an accuracy of 94.7% on data without sacrificing decentralization and resistant to collusion attacks in the network. Our results contribute to the design of reliable blockchain oracle infrastructure that decentralized finance (DeFi) and Web3 applications will rely on.
Ningthoujam et al. (Mon,) studied this question.