Towards Optimized Peer Connectivity in Blockchain Networks using Digital Twin

Optimized peer connectivity is crucial for reducing network latency in peer-to-peer (P2P) blockchain networks which in-turn contributes to achieve higher transaction rates. A new peer joins the P2P network by connecting to a random set of existing peers making these networks inherently random, resulting in increased netowrk latency. Connecting peers that are geographically proximal (nearest neighbor connectivity (NNC) approach) helps to reduce the network latency, however, it creates network imbalance and leads to formation of hub nodes. Therefore, having a global view of the entire network is essential for making informed decisions about peer connectivity while balancing the network. This paper introduces ‘Digital Twin for blockchain P2P Network’ (DTPN) that stores the real time information of the peers including their geo-location, connectivity and latencies. Further, this work proposes 'Proximal and Degree Balanced Connectivity Algorithm (PDBCA)‘ which leverages DTPN to strategically identify peers based on their geographical proximity and existing connection load for every new peer joining the P2P network. Experimental evaluations demonstrate that PDBCA improves overall network latency while balancing the P2P network compared to Random and NNC approaches.