ABSTRACT With the advent of blockchain technology, the need to employ third parties, known as oracles, to assist smart contracts has become clear. Blockchain smart contracts can only operate on on‐chain data and face the major challenge of not being able to communicate with the outside world. Blockchain oracles resolve this problem by bridging the gap between on‐chain and off‐chain data. Oracles can exhibit inappropriate behaviours or selfishly hide their true resources to maximise profits. Current research presents oracles as trusted entities without providing a robust evaluation mechanism for them. This type of design carries the risk of turning oracles into central points of failure. In this paper, a new approach inspired by reinforcement learning (Q‐learning) is proposed. We investigate the performance of oracles and select reliable and cost‐effective oracles from the blockchain network and compare the efficiency of the proposed method with previous well‐known methods. This method, which we call IPQL ‘Integer‐based pseudo‐Q‐Learning’, helps to identify reliable and cost‐effective oracles more effectively. The proposed model analyses the behaviour of oracles using a performance‐cost table and a framework inspired by the Q‐learning algorithm. The IPQL model allows all blockchain validators to verify the results obtained during the execution of the algorithm, thus addressing the issue of randomness in the limited structure of the blockchain. Finally, IPQL is compared with current state‐of‐the‐art methods based on accuracy, cost, and execution time. The results show that the proposed method leads to a 31% increase in success rate and 10% request redirection to safe oracles, a 22% reduction in request redirection to malicious oracles, and a 9.1% improvement in time.
Soleimany et al. (Thu,) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: