SIM: Achieving High Profit Through Integration of Selfish Strategy Into Innocent Mining

Abstract

Selfish mining, one of the most renowned attack in Bitcoin, involves a selfish miner withholding discovered blocks and broadcasting them at an opportune moment to gain higher rewards than honest mining. However, selfish mining and its variants rely on two assumptions: the attacker solely engages in infiltration mining within the victim pool (attack assumption) and the system operates in a perfect network environment (network assumption). In this paper, we propose a novel attack called Selfish in Innocent Mining (SIM). SIM expands the range of attacker’s behaviors by incorporating selfish mining into the traditional framework of innocent and infiltration mining, without increasing the attacker’s computational power. Initially, we analyze all possible states of chains in the system and their transition probabilities in the context of the SIM attack using Markov Chain. We determine the attacker’s rewards in one victim pool, multiple victim pools, and the miner’s dilemma within different cases. Subsequently, we examine the impact of an imperfect network environment on the attacker’s rewards within the SIM framework, focusing on the influence of unintentional fork rates on rewards. Our quantitative analysis demonstrates that the attacker’s rewards in SIM exceed those in power-adjusting withholding (PAW) by $1.9\times $ and $2.7\times $ in different network environments, respectively. The attacker’s rewards threshold reduced to 12.38% compared to other benchmarks.