Matching-Gossip: Optimizing Blockchain Broadcast Performance to Address the CAP Trilemma

Blockchain is developing rapidly, showing great application potential, but it also faces many challenges. Facing the challenge of CAP (Consistency, Availability, Partition tolerance) trilemma, the proposal of GBT-CHAIN (General Blockchain Trilemma-solved Blockchain) framework provides a theoretical solution. Although the GBT-CHAIN framework shows excellent theoretical performance, the actual test results still have some deviation from the theoretical value, because the network layer has become a new bottleneck when the physical layer and the consensus layer are optimized. Therefore, this paper proposes a new broadcast protocol: Matching-Gossip. Its direct neighbor node discovery mechanism and state broadcast mechanism reduce link duplication and node duplication, solve the topological mismatch problem of Gossip, and break through the performance bottleneck of network layer. According to the comparative analysis of small-scale simulation experiments and large-scale real-world experiments, Matching-Gossip reduces the convergence time, greatly reduces the network load, shows good adaptability to the physical layer topology in the CAP trilemma theoretical solution, and greatly improves the communication efficiency of the blockchain network. Matching-Gossip optimizes the broadcast method of the blockchain network layer and plays its linkage role in solving the CAP trilemma.