Improved gossip protocol for blockchain applications
Blockchain is a distributed digital ledger consisting of sequence of blocks. These blocks are groups of transactions that make up a sequence where each block contains the cryptographic hash of its prior block. To maintain huge networks, data integrity, and consistency some blockchain deployments uses a distributed communication method known as gossip protocol. In gossip protocol, every node/system relays any recent data to all its neighbors in an atomic fashion where each neighbor imitates the process until the whole system is synchronized. The number of iterations required to synchronize N hosts is O (log N) despite the failing nodes and data losses in transmission. However, the protocol is not cost effective when it comes to number of state exchanges and the transmission delay due to the redundant state information exchanges. To make gossip protocol more efficient; first, we propose a fail-proof algorithm that can be used under zero node failure scenarios; secondly, an opportunistic algorithm giving a rational, stable, and predictable correction phase and lastly a checking algorithm that obtains an iterative correction until a convergence or a synchronization across all the nodes in the system are reached. We implement these improvements and carefully analyze and demonstrate our findings.
Blockchain, Convergence, Cryptocurrency, Gossip protocol, Synchronization
Gokay Saldamli, Charit Upadhyay, Devika Jadhav, Rohit Shrishrimal, Bapugouda Patil, and Lo’ai Tawalbeh. "Improved gossip protocol for blockchain applications" Cluster Computing (2022): 1915-1926. https://doi.org/10.1007/s10586-021-03504-z