The Annals of Applied Probability

Uniform observability of hidden Markov models and filter stability for unstable signals

Ramon van Handel

Full-text: Open access


A hidden Markov model is called observable if distinct initial laws give rise to distinct laws of the observation process. Observability implies stability of the nonlinear filter when the signal process is tight, but this need not be the case when the signal process is unstable. This paper introduces a stronger notion of uniform observability which guarantees stability of the nonlinear filter in the absence of stability assumptions on the signal. By developing certain uniform approximation properties of convolution operators, we subsequently demonstrate that the uniform observability condition is satisfied for various classes of filtering models with white-noise type observations. This includes the case of observable linear Gaussian filtering models, so that standard results on stability of the Kalman–Bucy filter are obtained as a special case.

Article information

Ann. Appl. Probab., Volume 19, Number 3 (2009), 1172-1199.

First available in Project Euclid: 15 June 2009

Permanent link to this document

Digital Object Identifier

Mathematical Reviews number (MathSciNet)

Zentralblatt MATH identifier

Primary: 93E11: Filtering [See also 60G35]
Secondary: 60J25: Continuous-time Markov processes on general state spaces 62M20: Prediction [See also 60G25]; filtering [See also 60G35, 93E10, 93E11] 93B07: Observability 93E15: Stochastic stability

Nonlinear filtering prediction asymptotic stability observability hidden Markov models uniform approximation merging of probability measures


van Handel, Ramon. Uniform observability of hidden Markov models and filter stability for unstable signals. Ann. Appl. Probab. 19 (2009), no. 3, 1172--1199. doi:10.1214/08-AAP576.

Export citation


  • [1] Aliprantis, C. D. and Burkinshaw, O. (2006). Positive Operators. Springer, Dordrecht.
  • [2] Blackwell, D. and Dubins, L. (1962). Merging of opinions with increasing information. Ann. Math. Statist. 33 882–886.
  • [3] Budhiraja, A. and Ocone, D. (1997). Exponential stability of discrete-time filters for bounded observation noise. Systems Control Lett. 30 185–193.
  • [4] Budhiraja, A. and Ocone, D. (1999). Exponential stability in discrete-time filtering for nonergodic signals. Stochastic Process. Appl. 82 245–257.
  • [5] Chigansky, P. and Liptser, R. (2006). On a role of predictor in the filtering stability. Electron. Comm. Probab. 11 129–140 (electronic).
  • [6] Çinlar, E. (1972). Markov additive processes. I, II. Z. Wahrsch. Verw. Gebiete 24 85–93.
  • [7] Cooper, J. B. and Schachermayer, W. (1981). Uniform measures and co-Saks spaces. In Functional Analysis, Holomorphy, and Approximation Theory (Proc. Sem., Univ. Fed. Rio de Janeiro, Rio de Janeiro, 1978). Lecture Notes in Mathematics 843 217–246. Springer, Berlin.
  • [8] Crisan, D. and Heine, K. (2008). Stability of the discrete time filter in terms of the tails of noise distributions. J. Lond. Math. Soc. 78 441–458.
  • [9] Crisan, D. and Rozovsky, B. (2009). The Oxford University Handbook of Nonlinear Filtering. Oxford Univ. Press. To appear.
  • [10] D’Aristotile, A., Diaconis, P. and Freedman, D. (1988). On merging of probabilities. Sankhyā Ser. A 50 363–380.
  • [11] Diaconis, P. and Freedman, D. (1986). On the consistency of Bayes estimates. Ann. Statist. 14 1–67.
  • [12] Dudley, R. M. (1966). Convergence of Baire measures. Studia Math. 27 251–268.
  • [13] Dudley, R. M. (2002). Real Analysis and Probability. Cambridge Studies in Advanced Mathematics 74. Cambridge Univ. Press, Cambridge.
  • [14] Edwards, R. E. (1966). Uniform approximation on noncompact spaces. Trans. Amer. Math. Soc. 122 249–276.
  • [15] Ethier, S. N. and Kurtz, T. G. (1986). Markov Processes: Characterization and Convergence. Wiley, New York.
  • [16] Folland, G. B. (1999). Real Analysis: Modern Techniques and Their Applications, 2nd ed. Wiley, New York.
  • [17] Kurtz, T. G. (1998). Martingale problems for conditional distributions of Markov processes. Electron. J. Probab. 3 29.
  • [18] LeGland, F. and Oudjane, N. (2003). A robustification approach to stability and to uniform particle approximation of nonlinear filters: The example of pseudo-mixing signals. Stochastic Process. Appl. 106 279–316.
  • [19] Ocone, D. and Pardoux, E. (1996). Asymptotic stability of the optimal filter with respect to its initial condition. SIAM J. Control Optim. 34 226–243.
  • [20] Orstein, D. and Sucheston, L. (1970). An operator theorem on L1 convergence to zero with applications to Markov kernels. Ann. Math. Statist. 41 1631–1639.
  • [21] Oudjane, N. and Rubenthaler, S. (2005). Stability and uniform particle approximation of nonlinear filters in case of nonergodic signals. Stoch. Anal. Appl. 23 421–448.
  • [22] Pachl, J. K. (1979). Measures as functionals on uniformly continuous functions. Pacific J. Math. 82 515–521.
  • [23] Parthasarathy, K. R. (1967). Probability Measures on Metric Spaces. Probability and Mathematical Statistics 3. Academic Press, New York.
  • [24] Revuz, D. and Yor, M. (1999). Continuous Martingales and Brownian Motion, 3rd ed. Grundlehren der Mathematischen Wissenschaften [Fundamental Principles of Mathematical Sciences] 293. Springer, Berlin.
  • [25] Rudin, W. (1991). Functional Analysis, 2nd ed. McGraw-Hill Inc., New York.
  • [26] Stannat, W. (2005). Stability of the filter equation for a time-dependent signal on ℝd. Appl. Math. Optim. 52 39–71.
  • [27] Tsukahara, H. (1999). A limit theorem for the prediction process under absolute continuity. In Séminaire de Probabilités, XXXIII. Lecture Notes in Mathematics 1709 397–404. Springer, Berlin.
  • [28] van Handel, R. (2009). The stability of conditional Markov processes and Markov chains in random environments. Ann. Probab. To appear.
  • [29] van Handel, R. (2009). Observability and nonlinear filtering. Probab. Theory Related Fields. To appear.
  • [30] Yor, M. (1977). Sur les théories du filtrage et de la prédiction. In Séminaire de Probabilités, XI (Univ. Strasbourg, Strasbourg, 1975/1976). Lecture Notes in Mathematics 581 257–297. Springer, Berlin.