Towards Overcoming the Undercutting Problem

For Bitcoin and similar cryptocurrencies, their mining processes are currently incentivized with fixed block rewards and voluntary transaction fees. However, the block rewards are supposed to vanish gradually and the remaining incentive of transaction fees is optional and arbitrary. Under those circumstances, Carlsten et al.[CCS~2016] find that an interesting undercutting attack, where the attacker deliberately forks an existing chain by leaving wealthy transactions unclaimed to attract petty complaint miners to its fork, can become the equilibrium strategy for miners. Motivated by similar phenomenons in economics, we take a closer look at the undercutting analysis and find the result to be questionable: In [CCS~2016], fees are accumulated at a fixed rate and miners can collect all unclaimed fees regardless of block size limit, which is often not feasible in practice. Besides, ignoring a potentially large amount of fees unclaimable in a single block can inaccurately inflate the profitability of undercutting. In this work, we define a model that considers claimable fees based on available transactions that can be assembled into the block size limit and upgrades petty compliant miners to be rational where they decide whether to move to other chains subject to expected returns from different choices. In this new model, we first identify the conditions that are necessary to make undercutting profitable. Second, we propose a defense against undercutting by manipulating transactions selected into the new block to invalidate the above-identified conditions. Finally, we complement the above analytical results with an experimental analysis over Bitcoin and Monero. We demonstrate that our conditions for undercutting to be profitable are effective (an increase of 0.5-4.5% in Bitcoin and 8% in Monero) and the avoidance technique fulfills its purpose of allowing miners to earn around fair shares.