Abstract: Selfish mining is a mining strategy in the blockchain, which can increase profits by selectively publishing the “blocks” dug up. This paper proposes SAPV model based on probability distribution, aiming at the problems that selfish mining destroys the normal mining process and wastes the mining power. The typical attacks of blockchain are summarized firstly, then the process of selfishness minings is analyzed, and the different states of selfish mining are gottern, the probability distribution of selfish mining is calculated in different situation lastly. A new method of comparing the size of the probability of selfish and honesty mining to determine whether the hidden blocks are published is given, which optimizes the relative revenue. The experiment simulates the relative revenue share of the selfish mining pool under different mining power, and this paper analyzes the impact of different computing power on the reward of the pool. Lastly the threshold of selfish mining is proposed for ensuring the safe processing of mining. The paper provides the ideas to solve the problem of blockchain security for prospective research.

Key words: selfish mining, blockchain, mining blocks, probability distribution, pool revenue

摘要： 自私挖掘是区块链中的一种挖掘策略，通过选择性地公布挖到的区块以“增加”自己的收益。针对目前区块链挖掘中的自私挖掘行为破坏正常挖掘过程、浪费算力的问题，给出了基于概率的SAPV决策模型。首先总结了针对区块链的典型攻击，详细分析自私挖掘的过程，得到自私挖掘过程中出现的不同状态，计算了自私挖掘在不同状态下的概率分布。为增大自私挖掘的相对收益份额，给出了通过求解自私与诚实挖掘概率大小的方法来决定是否公布隐藏的区块。实验模拟了不同算力下自私挖掘池的相对收益，分析了不同算力对矿池收益的影响，最后给出了保证系统诚实节点安全运行的算力阈值，为区块链安全的进一步研究提供参考价值。

关键词: 自私挖掘, 区块链, 区块挖掘, 概率分布, 矿池收益