Decision diagrams (DDs) are widely used in system verification to compute and store the state space of finite discrete events dynamic systems (DEDSs). DDs are organized into levels, and it is well known that the size of a DD encoding a given set may be very sensitive to the order in which the variables capturing the state of the system are mapped to levels. Computing optimal orders is NP-hard. Several heuristics for variable order computation have been proposed, and metrics have been introduced to evaluate these orders. However, we know of no published evaluation that compares the actual predictive power for all these metrics. We propose and apply a methodology to carry out such an evaluation, based on the correlation between the metric value of a variable order and the size of the DD generated with that order. We compute correlations for several metrics from the literature, applied to many variable orders built using different approaches, for 40 DEDSs taken from the literature. Our experiments show that these metrics have correlations ranging from “very weak or nonexisting” to “strong.” We show the importance of highly correlating metrics on variable order heuristics, by defining and evaluating two new heuristics (an improvement of the well-known Force heuristic and a metric-based simulated annealing), as well as a meta-heuristic (that uses a metric to select the “best” variable order among a set of different variable orders).
Variable order metrics for decision diagrams in system verification
Amparore, Elvio G.;Donatelli, Susanna;
2020-01-01
Abstract
Decision diagrams (DDs) are widely used in system verification to compute and store the state space of finite discrete events dynamic systems (DEDSs). DDs are organized into levels, and it is well known that the size of a DD encoding a given set may be very sensitive to the order in which the variables capturing the state of the system are mapped to levels. Computing optimal orders is NP-hard. Several heuristics for variable order computation have been proposed, and metrics have been introduced to evaluate these orders. However, we know of no published evaluation that compares the actual predictive power for all these metrics. We propose and apply a methodology to carry out such an evaluation, based on the correlation between the metric value of a variable order and the size of the DD generated with that order. We compute correlations for several metrics from the literature, applied to many variable orders built using different approaches, for 40 DEDSs taken from the literature. Our experiments show that these metrics have correlations ranging from “very weak or nonexisting” to “strong.” We show the importance of highly correlating metrics on variable order heuristics, by defining and evaluating two new heuristics (an improvement of the well-known Force heuristic and a metric-based simulated annealing), as well as a meta-heuristic (that uses a metric to select the “best” variable order among a set of different variable orders).File | Dimensione | Formato | |
---|---|---|---|
Variable order metrics for decision diagrams in system verification - STTT.pdf
Accesso aperto
Tipo di file:
POSTPRINT (VERSIONE FINALE DELL’AUTORE)
Dimensione
2.48 MB
Formato
Adobe PDF
|
2.48 MB | Adobe PDF | Visualizza/Apri |
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.