Mean-Field Analysis of Markov Models with Reward Feedback
We extend the population continuous time Markov chain formalism so that the state space is augmented with continuous variables accumulated over time as functions of component populations. System feedback can be expressed using accumulations that in turn can influence the Markov chain behaviour via functional transition rates. We show how to obtain mean-field differential equations capturing means and higher-order moments of the discrete populations and continuous accumulation variables. We also provide first- and second-order convergence results and suggest a novel normal moment closure that can greatly improve the accuracy of means and higher moments.
We demonstrate how such a framework is suitable for modelling feedback from globally-accumulated quantities such as energy consumption, cost or temperature. Finally, we present a worked example modelling a hypothetical heterogeneous computing cluster and its interaction with air conditioning units.
KeywordsMarkov Model Process Algebra Discrete Population Discrete State Space Component Population
Unable to display preview. Download preview PDF.
- 1.Billingsley, P.: Convergence of Probability Measures. John Wiley & Sons (1968)Google Scholar
- 2.Bortolussi, L.: Hybrid Limits of Continuous Time Markov Chains. In: QEST 2011, pp. 3–12. IEEE, Aachen (2011)Google Scholar
- 5.Davis, M.H.A.: Markov models and optimization. Chapman & Hall/CRC (1993)Google Scholar
- 6.Ethier, S.N., Kurtz, T.G.: Markov Processes: Characterization and Convergence. Wiley (2005)Google Scholar
- 7.Gast, N., Bruno, G.: A mean field model of work stealing in large-scale systems. In: SIGMETRICS, p. 13. ACM, New York (2010)Google Scholar
- 12.Hillston, J.: Fluid flow approximation of PEPA models. In: QEST, pp. 33–42. IEEE (September 2005)Google Scholar
- 14.Kallenberg, O.: Foundations of Modern Probability. Springer (2002)Google Scholar
- 15.Khadim, U.: A comparative study of process algebras for hybrid systems. Computer Science Report 06–23, Technische Universiteit Eindhoven (2006)Google Scholar
- 16.Klebaner, F.C.: Introduction to stochastic calculus with applications, 2nd edn. Imperial College Press (2006)Google Scholar
- 17.Rawson, A., Pfleuger, J., Cader, T.: Data Center Power Efficiency Metrics: PUE and DCiE. The Green Grid (2007)Google Scholar
- 18.de Souza e Silva, E., Gail, R.: An algorithm to calculate transient distributions of cummulative rate and impulse-based rewards. Communications in Statistics: Stochastic Models 14(3), 509–536 (1998)Google Scholar
- 20.Stefanek, A., Hayden, R.A., Bradley, J.T.: A new tool for the performance analysis of massively parallel computer systems. In: QAPL. Electronic Proceedings in Theoretical Computer Science (March 2010)Google Scholar
- 21.Stefanek, A., Hayden, R.A., Bradley, J.T.: Fluid Analysis of Energy Consumption using Rewards in Massively Parallel Markov Models. In: Computing, p. 121. ACM Press (2011)Google Scholar
- 22.Stefanek, A., Hayden, R.A., Bradley, J.T.: GPA – A Tool for Fluid Scalability Analysis of Massively Parallel Systems. In: QEST, pp. 147–148. IEEE (September 2011)Google Scholar
- 24.Telek, M., Rácz, S.: Numerical analysis of large Markovian reward models. Performance Evaluation, 36–37, 95–114 (August 1999)Google Scholar
- 26.Whitt, W.: Internet supplement to Stochastic-Process Limits (2002), http://www.columbia.edu/~ww2040/supplement.html