Analysis of Blockchain Forking on an Ethereum Network

Blockchain systems are on the rise, especially after the introduction and popularization of Bitcoin. The potential of the technology has been expanded and rebooted by Ethereum to support a broader range of applications through smart contracts. Transactions and contracts are activated every day in the Ethereum network and blocks are created at a very high rate. Often blocks are mined at conflicting times, which causes a split on the chain or a fork. Forks pose an inconsistency problem on the network until they are solved and the network agrees on the same instance of the blockchain. Transactions included on side forks, once the fork is resolved, are discarded and have to be mined again. Fork rates also impact the overall performance of the blockchain and the security of the system. This work studies how one of the main networking factors, namely communication delay, impacts forking of the blockchain. An analytical model for forking probability is presented. Furthermore, the impact of block distribution delays on the amount of forks is quantified using a private Ethereum blockchain in controlled experiments in a lab.