Peregrine 2.0: Explaining Correctness of Population Protocols Through Stage Graphs

Population Protocols: Beyond Runtime Analysis

“…This time, thanks to new progress by Leroux on the theory of Petri nets, we proved that the correctness problem is decidable, although as hard as the reachability problem for Petri nets [31]. This was the starting point of a research program devoted to the theory and practice of verifying population protocols, which reached an important milestone in 2020 with the release of Peregrine 2.0, a verifier based on new theoretical results [15,32]. The first part of this note surveys this research, adding all the work carried out since 2017 to a brief previous survey [28].…”

“…We have developed a practical approach to the computation of stage graphs, implemented in the tool Peregrine 2.0 [32]. The design of the tool is guided by the theoretical results on stage graphs, and by the notion of dead transitions.…”

Black-box Testing Liveness Properties of Partially Observable Stochastic Systems

“…Second, we repeat the experiments of [6] on models from the standard PRISM Benchmark Suite [10] using our blackbox strategies. Finally, we conduct experiments on population protocols from the benchmark suite of the Peregrine tool [4,5]. Synthetic Experiments.…”

“…The specification states that for each initial configuration the agents eventually reach the right consensus (property holds/does not hold). We have tested our strategies on several protocols from the benchmark suite of Peregrine, the state-of-the-art model checker for population protocols [4,5]. The first protocol of Table 2 is faulty, but Peregrine cannot prove it; our strategy finds initial configurations for which the protocol exhibits a fault.…”