MRMSolve: Distribution Estimation of Large Markov Reward Models

MRMSolve is an analysis tool developed for the evaluation of large Markov Reward Models (MRM). The previous version of MRMSolve [8] provided only the moments of MRMs at arbitrary transient instant of time. This paper presents a new version of MRMSolve with new analysis features and software environment. The most important new functionality of MRMSolve is that it also makes distribution estimation of MRMs. MRMSolve can estimate the distribution of reward measures up to models with ∼106 states, and to the best of our knowledge no other algorithm can handle MRMs with more than ∼104 states.