Analyzing the one dimensional Ising model by probabilistic model checking

  • Authors:
  • Toshifusa Sekizawa;Tatsuhiro Tsuchiya;Tohru Kikuno;Koichi Takahashi

  • Affiliations:
  • Research Center for Verification and Semantics (CVS) and Osaka University;Osaka University;Osaka University;Research Center for Verification and Semantics (CVS)

  • Venue:
  • AsiaMS '07 Proceedings of the IASTED Asian Conference on Modelling and Simulation
  • Year:
  • 2007

Quantified Score

Hi-index 0.00

Visualization

Abstract

Probabilistic model checking is an emerging verification technology for probabilistic analysis. It has started being used not only in the computer science field but also in interdisciplinary fields. In this paper, we show that probabilistic model checking allows one to analyze magnetic behaviours of the one dimensional Ising model which describes physical phenomena of magnets. The Ising model consists of elementary microscopic objects called spins and is often statically analyzed using the Metropolis method. Spins are microscopic elements, while the interesting properties to be measured are macroscopic physical quantities such as the energy. To analyze the Ising model with probabilistic model checking, we build labelled Discrete Time Markov Chain (DTMC) models in which states represent spins and macroscopic physical quantities and transitions occur according to the Metropolis method. Two representative physical quantities, energy and magnetization, are focused on. We use PRISM, a probabilistic model checker, to assess these quantities. To this end, we provide specifications expressed in PCTL (Probabilistic real time Computation Tree Logic) to instruct PRISM to compute the quantities.