Research On The Mechanism Of Selfish Mining Under Repeated Game

The Proof of Work(PoW)is typically used in blockchain as a consensus mechanism.With Po W,the recording right and reward of the newly mined block are reasonably assigned through the mining competition among mining pools.Featured by simplicity to implementation and a high degree of decentralization,Po W has recently become the most accepted solution to reach consensus in a great number of blockchain projects.Nevertheless,the security of Po W relies on an essential assumption,that is,all mining pools perform honestly by strictly complying with the rules of Po W.Most recently,however,intensive research has revealed that it seems plausible to carry out a strategy of the so-called selfish mining in blockchain.The selfish mining indicates a mining pool is likely to hide its newly mined block to obtain mining advantage and reveals it at a proper time as a kind of competitive benefit(excess earning)is reaped.Obviously,the behavior of selfish mining greatly threatens the security of sustainable construction of blockchain and consequently poses extensive damage to the sustainable development of the blockchain ecosystem.Potential solutions to cope with selfish mining are based on a thorough understanding of its mechanism.However,more efforts have still been required to construct or reveal the fundamental model,long-term mechanism,and actual effectiveness of selfish mining.Firstly,existing selfish mining models fail to accurately characterize the actual running of blockchain due to unrealistic assumptions and incomplete analysis on factors of mining strategy,especially the ignorance of the impact of interactions between mining pools on their strategic choices.Secondly,existing research is not concerned about the salient fact that some mining pools strive for long-term profits at the expense of sacrificing shortterm profits during the sustainable construction of blockchain.As a result,the mechanism of selfish mining under the orientation of long-term optimum is still unclear.Thirdly,there is still a lack of empirical research based on actual blockchain projects to make a confirmation though there is extensive research that argues the existence of selfish mining,which shows that whether selfish mining mentioned in relevant research is effective in the actual scenario is still unclear.To circumvent the aforementioned issues,a model that characterizes the practical mining behaviors of mining pools and their interactions was firstly formulated;also,the mechanism of selfish mining under repeated game was conducted;besides,the actual effectiveness of relevant strategies was especially verified with real industrial blockchain data.The main contributions of the current thesis are listed as follows:Firstly,we propose a more precise selfish mining model to overcome the shortcomings of existing models.Concretely,based on our analysis of the source code and documents about the mainstream and open-source blockchain projects,we capture behavioral models of mining pools during the mining process from application models and subsequently propose the precise selfish mining model.The impact of those unrealistic assumptions or incomplete analyses on the research about the mechanism of mining strategy can be avoided in this proposed model.Secondly,we propose a kind of selfish mining that can achieve the goal of long-term optimum for the lack of research on the long-term mechanism of selfish mining.Based on the previous selfish mining model,we convert the sustainable competition between mining pools as a repeated game and then calculate the change of mining pools' profit when selecting different strategies.Experiments show that selfish mining is not the optimal strategy for the mining pool under the repeated game.Subsequently,we propose the customized ?-SM strategy for different mining pools.Compared with honest mining and existing selfish mining,the ?-SM strategy enables the mining pool under the repeated game to achieve a goal of long-term optimum.Simultaneously,the combination of ?-SM strategies selected by all mining pools is the equilibrium of the repeated game.Thirdly,aiming at the lack of research on the actual effectiveness of selfish mining,we conduct the empirical research to verify the actual effectiveness of the ?-SM strategy based on historical data from the actual blockchain project.Experiments show that some mining behaviors are deviating from the rules prescribed during the construction of blockchain,and behaviors of mining pools with different computing power are consistent with the ?-SM strategy which proves that the ?-SM strategy is more in line with the actual running of the blockchain.