The analysis presents a fully automated framework for evaluating selfish mining attacks in blockchain systems. By using Markov Decision Processes (MDPs), it simplifies a complex challenge into manageable components, offering formal correctness guarantees on the results. The flexible nature of the analysis allows for adjustments in system parameters, demonstrating how these changes impact the adversary's expected revenue. This adaptability is crucial for understanding the dynamics of attacks in varying conditions, making the analysis valuable for both researchers and practitioners in blockchain security.
Our analysis is agnostic to the values of system model and attack parameters and it is flexible to their changes. Hence, it allows us to tweak the parameter values and study their impact on the optimal expected relative revenue, while preserving formal guarantees on the correctness.
By modelling our selfish mining attack as an MDP and reducing the analysis to solving mean-payoff MDPs, we leverage existing methods for formal analysis of MDPs to obtain a fully automated analysis procedure, thus avoiding the necessity for tedious manual analyses.
Our analysis provides formal guarantees on the correctness of its output. Again, this is achieved by formally reducing our problem to solving mean-payoff MDPs for which exact algorithms with formal correctness guarantees are available.
Manual analysis of optimal selfish mining attacks is already challenging and technically involved for PoW blockchains, where the adversary only grows a single private fork.
#selfish-mining #blockchain-security #automated-analysis #markov-decision-processes #formal-verification
Collection
[
|
...
]