Out-of-Core Solution of Large Linear Systems of Equations Arising from Stochastic Modelling

Many physical or computer systems can be modelled as Markov chains. A range of solution techniques exist to address the state-space explosion problem, encountered while analysing such Markov models. However, numerical solution of these Markov chains is impeded by the need to store the probability vector(s) explicitly in the main memory. In this paper, we extend the earlier out-of-core methods for the numerical solution of large Markov chains and introduce an algorithm which uses a disk to hold the probability vector as well as the matrix. We give experimental results of the implementation of the algorithm for a Kanban manufacturing system and a flexible manufacturing system. Next, we describe how the algorithm can be modified to exploit sparsity structure of a model, leading to better performance. We discuss two models, a cyclic server polling system and a workstation cluster system, in this context and present results for the polling models. We also introduce a new sparse matrix storage format which can provide 30% or more saving over conventional schemes.

[1]  William H. Sanders,et al.  An Efficient Disk-Based Tool for Solving Large Markov Models , 1998, Perform. Evaluation.

[2]  Gianfranco Ciardo,et al.  A data structure for the efficient Kronecker solution of GSPNs , 1999, Proceedings 8th International Workshop on Petri Nets and Performance Models (Cat. No.PR00331).

[3]  Andrew S. Miner Efficient solution of GSPNs using canonical matrix diagrams , 2001, Proceedings 9th International Workshop on Petri Nets and Performance Models.

[4]  Gianfranco Ciardo What a Structural World , 2001 .

[5]  Sivan Toledo,et al.  A survey of out-of-core algorithms in numerical linear algebra , 1999, External Memory Algorithms.

[6]  David A. Patterson,et al.  Computer Architecture: A Quantitative Approach , 1969 .

[7]  Masahiro Fujita,et al.  Multi-Terminal Binary Decision Diagrams: An Efficient Data Structure for Matrix Representation , 1997, Formal Methods Syst. Des..

[8]  Marta Z. Kwiatkowska,et al.  PRISM: Probabilistic Symbolic Model Checker , 2002, Computer Performance Evaluation / TOOLS.

[9]  Susanna Donatelli Superposed stochastic automata: a class of stochastic Petri nets amenable to parallel solution , 1991, Proceedings of the Fourth International Workshop on Petri Nets and Performance Models PNPM91.

[10]  I. Duff,et al.  Direct Methods for Sparse Matrices , 1987 .

[11]  Kishor S. Trivedi,et al.  Stochastic Petri Net Models of Polling Systems , 1990, IEEE J. Sel. Areas Commun..

[12]  William H. Sanders,et al.  An Efficient Disk-Based Tool for Solving Very Large Markov Models , 1997, Computer Performance Evaluation.

[13]  M. Siegle,et al.  Multi Terminal Binary Decision Diagrams to Represent and Analyse Continuous Time Markov Chains , 1999 .

[14]  Susanna Donatelli,et al.  Superposed Stochastic Automata: A Class of Stochastic Petri Nets with Parallel Solution and Distributed State Space , 1993, Perform. Evaluation.

[15]  Joost-Pieter Katoen,et al.  On the use of model checking techniques for dependability evaluation , 2000, Proceedings 19th IEEE Symposium on Reliable Distributed Systems SRDS-2000.

[16]  Brigitte Plateau On the stochastic structure of parallelism and synchronization models for distributed algorithms , 1985, SIGMETRICS 1985.

[17]  William H. Sanders,et al.  "On-the-Fly'' Solution Techniques for Stochastic Petri Nets and Extensions , 1998, IEEE Trans. Software Eng..

[18]  Markus Siegle,et al.  Advances in Model Representations , 2001, PAPM-PROBMIV.

[19]  Marta Z. Kwiatkowska,et al.  Probabilistic symbolic model checking with PRISM: a hybrid approach , 2004, International Journal on Software Tools for Technology Transfer.

[20]  Yousef Saad,et al.  Iterative methods for sparse linear systems , 2003 .

[21]  Enrico Macii,et al.  Markovian analysis of large finite state machines , 1996, IEEE Trans. Comput. Aided Des. Integr. Circuits Syst..

[22]  G. Ciardo,et al.  ON THE USE OF KRONECKER OPERATORS FOR THE SOLUTION OF GENERALIZED STOCHASTIC PETRI NETS , 1996 .

[23]  Kishor S. Trivedi,et al.  A Decomposition Approach for Stochastic Reward Net Models , 1993, Perform. Evaluation.

[24]  William J. Stewart,et al.  Introduction to the numerical solution of Markov Chains , 1994 .

[25]  David A. Patterson,et al.  Computer architecture (2nd ed.): a quantitative approach , 1996 .