Resampling methods for particle filtering: identical distribution, a new method, and comparable study

  • Tian-cheng Li
  • Gabriel Villarrubia
  • Shu-dong Sun
  • Juan M. Corchado
  • Javier Bajo


Resampling is a critical procedure that is of both theoretical and practical significance for efficient implementation of the particle filter. To gain an insight of the resampling process and the filter, this paper contributes in three further respects as a sequel to the tutorial (Li et al., 2015). First, identical distribution (ID) is established as a general principle for the resampling design, which requires the distribution of particles before and after resampling to be statistically identical. Three consistent metrics including the (symmetrical) Kullback-Leibler divergence, Kolmogorov-Smirnov statistic, and the sampling variance are introduced for assessment of the ID attribute of resampling, and a corresponding, qualitative ID analysis of representative resampling methods is given. Second, a novel resampling scheme that obtains the optimal ID attribute in the sense of minimum sampling variance is proposed. Third, more than a dozen typical resampling methods are compared via simulations in terms of sample size variation, sampling variance, computing speed, and estimation accuracy. These form a more comprehensive understanding of the algorithm, providing solid guidelines for either selection of existing resampling methods or new implementations.

Key words

Particle filter Resampling Kullback-Leibler divergence Kolmogorov-Smirnov statistic 

CLC number



Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. Adiprawita, W., Ahmad, A.S., Sembiring, J., et al., 2011. New resampling algorithm for particle filter localization for mobile robot with 3 ultrasonic sonar sensor. Proc. Int. Conf. on Electrical Engineering and Informatics, p.1–6. [doi:10.1109/ICEEI.2011.6021733]Google Scholar
  2. Arulampalam, M.S., Maskell, S., Gordon, N., et al., 2002. A tutorial on particle filters for online nonlinear/non-Gaussian Bayesian tracking. IEEE Trans. Signal Process., 50(2):174–188. [doi:10.1109/78.978374]CrossRefGoogle Scholar
  3. Bashi, A.S., Jilkov, V.P., Li, X.R., et al., 2003. Distributed implementations of particle filters. Proc. 6th Int. Conf. on Information Fusion, p.1164–1171. [doi:10.1109/ICIF. 2003.177369]Google Scholar
  4. Beskos, A., Crisan, D., Jasra, A., 2014. On the stability of sequential Monte Carlo methods in high dimensions. Ann. Appl. Probab., 24(4):1396–1445. [doi:10.1214/13-AAP951]zbMATHMathSciNetCrossRefGoogle Scholar
  5. Bolic, M., Djuric, P.M., Hong, S., 2003. New resampling algorithms for particle filters. Proc. IEEE Int. Conf. on Acoustics, Speech and Signal Processing, p.589–592. [doi:10.1109/ICASSP.2003.1202435] Cappé, O.Google Scholar
  6. Godsill, S.J., Moulines, E., 2007. An overview of existing methods and recent advances in sequential Monte Carlo. Proc. IEEE, 95(5):899–924. [doi:10.1109/ JPROC.2007.893250]CrossRefGoogle Scholar
  7. Chen, Y., Xie, J., Liu, J.S., 2005. Stopping-time resampling for sequential Monte Carlo methods. J. R. Stat. Soc. B, 67(2):199–217. [doi:10.1111/j.1467–9868.2005.00497x]zbMATHMathSciNetCrossRefGoogle Scholar
  8. Choe, G.M., Wang, T., Liu, F., et al., 2014. An advanced integrated framework for moving object tracking. J. Zhejiang Univ.-Sci. C (Comput. & Electron.), 15(10): 861–877. [doi:10.1631/jzus.C1400006]CrossRefGoogle Scholar
  9. Choe, G.M., Wang, T., Liu, F., et al., 2015. Particle filter with spline resampling and global transition model. IET Comput. Vis., 9(2):184–197. [doi:10.1049/iet-cvi.2014.0106]CrossRefGoogle Scholar
  10. Crisan, D., Doucet, A., 2002. A survey of convergence results on particle filtering methods for practitioners. IEEE Trans. Signal Process., 50(3):736–746. [doi:10.1109/ 78.984773]MathSciNetCrossRefGoogle Scholar
  11. Crisan, D., Lyons, T., 1999. A particle approximation of the solution of the Kushner-Stratonovitch equation. Probab. Theory Related Fields, 115(4):549–578. [doi:10.1007/ s004400050249]zbMATHMathSciNetCrossRefGoogle Scholar
  12. Crisan, D., Del Moral, P., Lyons, T., 1998. Discrete Filtering Using Branching and Interacting Particle Systems. Markov Process. Related Fields, 5(3):293–318.MathSciNetGoogle Scholar
  13. Das, S.K., Mazumdar, C., 2013. Priori-sensitive resampling particle filter for dynamic state estimation of UUVs. Proc. 8th Int. Workshop on Systems, Signal Processing and Their Applications, p.384–389. [doi:10.1109/ WoSSPA.2013.6602396] DelGoogle Scholar
  14. Moral, P., Hu, P., Wu, L., 2012. On the concentration properties of interacting particle processes. Found. Trends Mach. Learn., 3(3–4):225–389. [doi:10.1561/ 2200000026]Google Scholar
  15. Djuric, P.M., Miguez, J., 2010. Assessment of nonlinear dynamic models by Kolmogorov-Smirnov statistics. IEEE Trans. Signal Process., 58(10):5069–5079. [doi:10.1109/ TSP.2010.2053707]MathSciNetCrossRefGoogle Scholar
  16. Djuric, P.M., Kotecha, J.H., Zhang, J., et al., 2003. Particle filtering. IEEE Signal Process. Mag., 20(5):19–38. [doi:10.1109/MSP.2003.1236770]CrossRefGoogle Scholar
  17. Douc, R., Cappé, O., 2005. Comparison of resampling schemes for particle filtering. Proc. 4th Int. Symp. on Image and Signal Processing and Analysis, p.64–69. [doi:10.1109/ISPA.2005.195385]Google Scholar
  18. Douc, R., Moulines, E., Olsson, J., 2014. Long-term stability of sequential Monte Carlo methods under verifiable conditions. Ann. Appl. Probab., 24(5):1767–1802. [doi:10. 1214/13-AAP962]zbMATHMathSciNetCrossRefGoogle Scholar
  19. Doucet, A., de Freitas, N., Gordon, N., 2001. Sequential Monte Carlo Methods in Practice. Springer, New York, USA. [doi:10.1007/978–1-4757–3437-9]zbMATHCrossRefGoogle Scholar
  20. Efron, B., Rogosa, D., Tibshirani, R., 2015. Resampling methods of estimation. In: Wright, J.D. (Ed.), International Encyclopedia of the Social & Behavioral Sciences (2nd Ed.). Elsevier, Oxford, p.492–495. [doi:10.1016/B978–0-08–097086–8.42165–3]CrossRefGoogle Scholar
  21. Fearnhead, P., Clifford, P., 2003. On-line inference for hidden Markov models via particle filters. J. R. Stat. Soc. Ser. B, 65(4):887–899. [doi:10.1111/1467–9868.00421]zbMATHMathSciNetCrossRefGoogle Scholar
  22. Fearnhead, P., Liu, Z., 2007. On-line inference for multiple changepoint problems. J. R. Stat. Soc. Ser. B, 69(4): 589–605. [doi:10.1111/j.1467–9868.2007.00601x]MathSciNetCrossRefGoogle Scholar
  23. Fox, D., 2003. Adapting the sample size in particle filters through KLD-sampling. Int. J. Robot. Res., 22(12):985–1003. [doi:10.1177/0278364903022012001]CrossRefGoogle Scholar
  24. Godsill, S., Vermaak, J., Ng, W., et al., 2007. Models and algorithms for tracking of maneuvering objects using variable rate particle filters. Proc. IEEE, 95(5):925–952. [doi:10.1109/JPROC.2007.894708]CrossRefGoogle Scholar
  25. Gordon, N., Salmond, D., Smith, A.F.M., 1993. Novel approach to nonlinear/non-Gaussian Bayesian state estimation. IEE Proc. F, 140(2):107–113. [doi:10.1049/ip-f-2.1993.0015]Google Scholar
  26. Gustafsson, F., 2010. Particle filter theory and practice with positioning applications. IEEE Aeros. Electron. Syst. Mag., 25(7):53–82. [doi:10.1109/MAES.2010.5546308]CrossRefGoogle Scholar
  27. Hol, J.D., Schon, T.B., Gustafsson, F., 2006. On resampling algorithms for particle filters. Proc. IEEE Nonlinear Statistical Signal Processing Workshop, p.79–82. [doi:10.1109/NSSPW.2006.4378824]Google Scholar
  28. Hong, S., Shi, Z., Chen, J., et al., 2010. A low-power memory-efficient resampling architecture for particle filters. Circ. Syst. Signal Process., 29(1):155–167. [doi:10. 1007/s00034–009-9117–4]zbMATHCrossRefGoogle Scholar
  29. Hu, X.L., Schon, T.B., Ljung, L., 2011. A general convergence result for particle filtering. IEEE Trans. Signal Process., 59(7):3424–3429. [doi:10.1109/TSP.2011. 2135349]MathSciNetCrossRefGoogle Scholar
  30. Kalman, R.E., 1960. A new approach to linear filtering and prediction problems. J. Basic Eng., 82(1):35–45. [doi:10. 1115/1.3662552]CrossRefGoogle Scholar
  31. Kitagawa, G., 1996. Monte Carlo filter and smoother and non-Gaussian nonlinear state space models. J. Comput. Graph. Stat., 5(1):1–25. [doi:10.1080/10618600.1996. 10474692]MathSciNetGoogle Scholar
  32. Kong, A., Liu, J.S., Wong, W.H., 1994. Sequential imputations and Bayesian missing data problems. J. Am. Stat. Assoc., 89(425):278–288. [doi:10.1080/01621459.1994. 10476469]zbMATHCrossRefGoogle Scholar
  33. Kullback, S., Leibler, R.A., 1951. On information and sufficiency. Ann. Math. Stat., 22(1):79–86. [doi:10.1214/ aoms/1177729694]zbMATHMathSciNetCrossRefGoogle Scholar
  34. Kwak, N., Kim, G.W., Lee, B.H., 2008. A new compensation technique based on analysis of resampling process in FastSLAM. Robotica, 26(2):205–217. [doi:10.1017/S026 3574707003773]CrossRefGoogle Scholar
  35. Lang, H., Li, T., Villarrubia, G., et al., 2015. An adaptive particle filter for indoor robot localization. Proc. 6th Int. Symp. on Ambient Intelligence, p.45–55. [doi:10.1007/978–3-319–19695-4_5]Google Scholar
  36. Lenstra, H.W., 1983. Integer programming with a fixed number of variables. Math. Oper. Res., 8(4):538–548. [doi:10.1287/moor.8.4.538]zbMATHMathSciNetCrossRefGoogle Scholar
  37. Li, T., Sun, S., 2010. Double-resampling based Monte Carlo localization for mobile robot. Acta Autom. Sin., 36(9): 1279–1286. [doi:10.3724/SP.J.1004.2010.01279]CrossRefGoogle Scholar
  38. Li, T., Sattar, T.P., Sun, S., 2012. Deterministic resampling: unbiased sampling to avoid sample impoverishment in particle filters. Signal Process., 92(7):1637–1645. [doi:10.1016/jsigpro.2011.12.019]CrossRefGoogle Scholar
  39. Li, T., Sattar, T.P., Tang, D., 2013a. A fast resampling scheme for particle filters. Proc. Constantinides Int. Workshop on Signal Processing, p.1–4. [doi:10.1049/ ic.2013.0002]Google Scholar
  40. Li, T., Sun, S., Sattar, T.P., 2013b. Adapting sample size in particle filters through KLD-resampling. Electron. Lett., 46(2):740–742. [doi:10.1049/el.2013.0233]CrossRefGoogle Scholar
  41. Li, T., Sun, S., Sattar, T.P., et al., 2014. Fight sample degeneracy and impoverishment in particle filters: a review of intelligent approaches. Expert Syst. Appl., 41(8):3944–3954. [doi:10.1016/jeswa.2013.12.031]CrossRefGoogle Scholar
  42. Li, T., Bolic, M., Djuric, P.M., 2015. Resampling methods for particle filtering: classification, implementation, and strategies. IEEE Signal Process. Mag., 32(3):70–86. [doi:10.1109/MSP.2014.2330626]CrossRefGoogle Scholar
  43. Li, T., Sun, S., Bolic, M., et al., 2016. Algorithm design for parallel implementation of the SMC-PHD filter. Signal Process., 119:115–127. [doi:10.1016/jsigpro.2015.07.013]CrossRefGoogle Scholar
  44. Liu, J.S., Chen, R., 1998. Sequential Monte Carlo methods for dynamic systems. J. Am. Stat. Assoc., 93(443): 1032–1044. [doi:10.1080/01621459.1998.10473765]zbMATHCrossRefGoogle Scholar
  45. Liu, J.S., Chen, R., Logvinenko, T., 2001. A theoretical framework for sequential importance sampling and resampling. In: Doucet, A., de Freitas, N., Gordon, N. (Eds.), Sequential Monte Carlo Methods in Practice. Springer, USA, p.225–246. [doi:10.1007/978–1-4757–3437-9_11]CrossRefGoogle Scholar
  46. Mbalawata, I.S., Särkkä, S., 2016. Moment conditions for convergence of particle filters with unbounded importance weights. Signal Process., 118:133–138. [doi:10.1016/jsigpro.2015.06.018]CrossRefGoogle Scholar
  47. Míguez, J., Bugallo, M.F., Djuric, P.M., 2004. A new class of particle filters for random dynamical systems with unknown statistics. EURASIP J. Adv. Signal Process., 15:2278–2294. [doi:10.1155/S1110865704406039]CrossRefGoogle Scholar
  48. Morelande, M.R., Zhang, A.M., 2011. A mode preserving particle filter. Proc. IEEE Int. Conf. on Acoustics, Speech and Signal Processing, p.3984–3987. [doi:10. 1109/ICASSP.2011.5947225]Google Scholar
  49. Murray, L., 2012. GPU acceleration of the particle filter: the Metropolis resampler. arXiv:1202.6163v1.Google Scholar
  50. Nielsen, F., 2010. A family of statistical symmetric divergences based on Jensen’s inequality. arXiv:1009.4004.Google Scholar
  51. Pérez, C.J., Martín, J., Rufo, M.J., et al., 2005. Quasi-random sampling importance resampling. Commun. Stat. Simul. Comput., 34(1):97–112. [doi:10.1081/SAC-200047112]zbMATHCrossRefGoogle Scholar
  52. Robert, C.P., Casella, G., 1999. Monte Carlo Statistical Methods. Springer, New York. [doi:10.1007/978–1-4757–4145-2]zbMATHCrossRefGoogle Scholar
  53. Rubin, D.B., 1987. The calculation of posterior distribution by data augmentation: Comment: a noniterative sampling/ importance resampling alternative to the data augmentation algorithm for creating a few imputations when fractions of missing information are modest: the SIR algorithm. J. Am. Stat. Assoc., 82(398):543–546. [doi:10. 2307/2289460]Google Scholar
  54. Sileshi, B.G., Ferrer, C., Oliver, J., 2013. Particle filters and resampling techniques: importance in computational complexity analysis. Proc. Conf. on Design and Architectures for Signal and Image Processing, p.319–325.Google Scholar
  55. Simonetto, A., Keviczky, T., 2009. Recent developments in distributed particle filtering: towards fast and accurate algorithms. Proc. 1st IFAC Workshop on Estimation and Control of Networked Systems, p.138–143. [doi:10. 3182/20090924–3-IT-4005.00024]Google Scholar
  56. Stano, P.M., Lendek, Z., Babuška, R., 2013. Saturated particle filter: almost sure convergence and improved resampling. Automatica, 49(1):147–159. [doi:10.1016/j. automatica.2012.10.006]zbMATHMathSciNetCrossRefGoogle Scholar
  57. Sutharsan, S., Kirubarajan, T., Lang, T., et al., 2012. An optimization-based parallel particle filter for multitarget tracking. IEEE Trans. Aeros. Electron. Syst., 48(2):1601–1618. [doi:10.1109/TAES.2012.6178081]CrossRefGoogle Scholar
  58. Topsoe, F., 2000. Some inequalities for information divergence and related measures of discrimination. IEEE Trans. Inform. Theory, 46(4):1602–1609. [doi:10.1109/ 18.850703]MathSciNetCrossRefGoogle Scholar
  59. Wang, Y., Djuric, P.M., 2013. Sequential estimation of linear models in distributed settings. Proc. 21st European Signal Processing Conf., p.1–5.Google Scholar
  60. Whiteley, N., 2013. Stability properties of some particle filters. Ann. Appl. Probab., 23(6):2500–2537. [doi:10.1214/ 12-AAP909]zbMATHMathSciNetCrossRefGoogle Scholar
  61. Zhi, R., Li, T., Siyau, M.F., et al., 2014. Applied technology in adapting the number of particles while maintaining the diversity in the particle filter. Adv. Mater. Res., 951:202–207. [doi:10.4028/wwwscientificnet/AMR.951.202]CrossRefGoogle Scholar

Copyright information

© Journal of Zhejiang University Science Editorial Office and Springer-Verlag Berlin Heidelberg 2015

Authors and Affiliations

  • Tian-cheng Li
    • 1
    • 2
  • Gabriel Villarrubia
    • 1
  • Shu-dong Sun
    • 2
  • Juan M. Corchado
    • 1
    • 4
  • Javier Bajo
    • 3
  1. 1.BISITE Group, Faculty of ScienceUniversity of SalamancaSalamancaSpain
  2. 2.School of Mechanical EngineeringNorthwestern Polytechnical UniversityXi’anChina
  3. 3.Department of Artificial IntelligenceTechnical University of MadridMadridSpain
  4. 4.Osaka Institute of TechnologyAsahi-ku Ohmiya, OsakaJapan

Personalised recommendations