A Blockchain Network Communication Architecture Based on information-Centric Networking

Blockchain technology, as a distributed ledger technology, is becoming increasingly popular in various fields. However, the performance of blockchain limits its development in practical applications. The mismatch between traditional blockchain P2P networks and the underlying network topology can cause broadcast delay and network bandwidth problems, thereby affecting the performance of the blockchain. In this paper, we propose a blockchain network communication architecture called BLOCK-ICN based on Information-Centric Networking (ICN). This architecture supports the deployment of blockchain applications on ICN nodes with computing and storage capabilities, is compatible with existing P2P networks, and provides acceleration services for them. BLOCK-ICN first partitions the network through an enhanced resolution system, proposes a proxy node election algorithm based on the upstream and downstream relationships of ICN nodes within the domain, constructs an ICN blockchain network topology that matches the underlying network topology, and optimizes the keep-alive protocol between proxy nodes and domain nodes based on historical packet information between nodes. Finally, a parallel tree broadcasting protocol based on ICN multicast is proposed in the ICN blockchain network, where ICN blockchain nodes broadcast messages through multicast and propagate messages to P2P network neighboring nodes.To verify the effectiveness of this architecture, we have conducted simulation experiments of BLOCK-ICN based on the improvement and implementation of SimBlock. The experimental results show that, with a network node scale of 16,000 and a broadcast coverage of 90\%, compared with the bitcoin network, BLOCK-ICN ensures network scalability and reduces 84.9\% of the block broadcast delay and 68.8\% of the total network bandwidth consumption when the proportion of ICN nodes is 100\%. When the proportion of ICN nodes is 1\%, the broadcast delay of blocks to 90\% and 50\% of the whole network is reduced by 25\% and 33.2\%, respectively.