An In-Depth Look at Forking-Based Attacks in Ethereum With PoW Consensus
Soosan Naderi Mighan, Jelena Mišić, Vojislav B. Mišić, Xiaolin Chang
Abstract
In this paper, we analyze the performance of Ethereum data distribution network using a probabilistic model which allows accurate modeling of data propagation but also of forking, which happens when the blockchain maintained by the network temporarily splits into multiple versions due to a disagreement over the validity of a particular block. We also investigate the duration of inconsistent states of the ledger, which refers to the amount of time that the network remains split or partitioned. Finally, we model the block withholding attack and block slowdown attack, and analyze their impact on network performance of the network in terms of quality indicators such as block delivery time, the duration of ledger inconsistency, and forking probability. We also propose countermeasures for the block withholding attack.