Adaptive mining difficulty for blockchain to resist selfish mining attack

Aim: A blockchain provides data consistency and builds a fair mining environment for a network by using a consensus mechanism such as proof of work (PoW) and proof of stake. However, selfish mining is a well-known mining attack. It can reduce the fairness and destabilize the network, especially for a PoW-based blockchain. Therefore, in this paper, we propose a new approach, named the adaptive mining difficulty adjustment protocol, which can deter a selfish attack. Methods: We propose using the unit profit as an improved version of the relative revenue, because it is more flexible for calculating the miners’ profits in different periods and can be used to analyze repeated mining games. Based on the unit profit, we propose using the adaptive mining-difficulty adjustment protocol to reduce an attacker’s profit. Our protocol evaluates the effective hash power in a network more accurately and corrects the mining difficulty. Moreover, we introduce the discount factor and model the long-term profit to analyze the impact of a miner’s patience on its future profit. Results: We used an open-source simulator to simulate the competition between a selfish miner and an honest miner and determined their profits under different protocols. Our experimental results show that our protocol can effectively raise the attack threshold, but it reduces the total network profit if the attacker still attacks. However, the long-term profit model shows that a more patient attacker needs to invest more hash power and pay increased mining costs to maintain its attack. Conclusion: We conclude that, under our protocol, selfish attackers will tend to become honest miners if their hash power does not exceed the threshold, which means that our protocol can effectively deter selfish attacks and some variants of selfish attacks. Briefly, our protocol can correct the mining difficulty and leads to a more stable and fairer mining environment for a PoW-based blockchain.