DQN-Based Dynamic Query Optimization Method of Data Cross-Chain Based on Multi-channel Relay Chain

With the rapid development of consortium blockchain technology, the interaction between consortium blockchains has attracted more and more attention. However, most of the traditional cross-chain technologies focus on value cross-chain. There is still a lot of research space for data cross-chain mechanism. Therefore, this paper proposes a DQN-based dynamic query optimization method of data cross-chain based on multi-channel relay chain (mRDC-DQ). In order to improve the efficiency of information interaction, this method constructs a multi-channel relay alliance chain architecture to realize parallel processing, and proposes a relay alliance chain priority query mechanism, which uses the relay alliance chain as a buffer for high-frequency query data. And a query filter is set in the block header of the relay block to achieve efficient search. Reinforcement learning is combined with the cross-chain mechanism for the first time, and the DQN model is used to dynamically select the optimal parameters according to the real-time state of the cross-chain system, so that the resources are fully utilized and the best cross-chain query efficiency of data is achieved. Experiments show that the proposed scheme has higher information query and interaction efficiency than the existing cross-chain schemes.