Selfish Mining Simulator

Selfish mining literature has traditionally focused on MDP to probe for theoretically ideal strategies. However, these fail to integrate crucial factors like network latency, difficulty adjustment, and node/pool internal processing delays, which can significantly affect the outcome of a real-world attack. The Monero Simulator incorporates these elements, plus discrete asynchronous agentic mining pools, into a configurable Monte-Carlo framework capable of sweeping over a broad set of input parameters. Tunable parameters include: pools/hashrate, difficulty adjustment constants, network latency/bandwidth, and pool strategies. Additionally, we introduce a novel simplification whereby the set of selfish strategy state-diagrams can be modeled with just two, integer input parameters. Preliminary results corroborate the classic selfish mining model, while challenging the relative profitability of “stubborn” variants. Finally, the simulator offers a versatile environment for testing selfish mining countermeasures, and could be extended to probe other aspects of Monero’s design.