Answers are grounded in this paper and cite their sources.
Sharding is a promising approach to enhancing blockchain scalability, and it has recently attracted widespread attention. In order to ensure the security of sharding blockchains, many existing systems that utilize non-cooperative consensus randomly allocate a large number of nodes to each shard, which poses difficulties for deployment in real-world scenarios. Most of the current cooperative consensus protocols depend on a global consensus. This reliance introduces additional consensus steps and restricts their scalability. Furthermore, the communication overhead in these protocols is significant and their transaction processing parallelism is low. In this paper, we propose CoCSV, a decentralized cross-shard verification protocol consisting of three phases.Specifically, it features a VERIFY phase, during which shard leaders exchange their Commitment Intentions regarding blocks generated by other shards. To enhance the protocol performance, we then introduce Urd, a 6P-3PC protocol achieving a sixfold increase in parallelism. It allows a cross-shard transaction to commit in just one voting round on average. We provide a theoretical analysis to demonstrate that our protocol can achieve a high level of security when compared to existing non-cooperative protocols. Finally, we implement a prototype of Urd using Go and conduct experiments on up to 440 nodes. Our evaluation demonstrates that the proposed protocol increases the transaction throughput by 2.877 times when there are 11 shards and 440 nodes, while incurring only 9.6% more bandwidth overhead for cross-shard verification.
No takes yet. Share an insight, caveat, or question.
Dong et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: