Calibrating the performance and security of blockchains via information propagation delays: revisiting an old approach with a new perspective

Miners of a blockchain exchange information about blocks and transactions with one another via a peer-to-peer (P2P) network. The speed at which they learn of new blocks and transactions in the network determines the likelihood of forks in the chain, which in turn has implications for the efficiency as well as security of proof-of-work (PoW) blockchains. Despite the importance of information propagation delays in a blockchain's peer-to-peer network, little is known about them. The last known empirical study was conducted, for instance, by Decker and Wattenhofer in 2013 [11]. In this paper, we revisit the work of Decker and Wattenhofer on information propagation delays in Bitcoin. We update their measurement methodology to accommodate the changes made to the P2P network protocols since 2013. We also expand our measurement effort to include three other widely used blockchains, namely Bitcoin Cash, Litecoin, and Dogecoin. We reveal that block propagation delays have drastically reduced since 2013: The majority of peers in all four blockchains learn of a newly mined block within one second; the likelihood of forks is, consequently, low. Though blockchains networks have become quite efficient (i.e., have low delays), we observe that a significant number of nodes of these blockchains are present in cloud-provider networks and, more importantly, state-owned network providers; such deployments have crucial security implications for blockchains.