Cyclic Cramér-Rao-type bounds for periodic parameter estimation

Cyclic lower bounds establish fundamental limits in parametric statistical models with a periodic nature. Non-Bayesian lower bounds on the mean-cyclic-error (MCE) of any cyclic-unbiased estimator, in the Lehmann sense, are useful for performance analysis and system design in periodic estimation. In this paper, we derive two cyclic versions of the Cramér-Rao bound (CRB) based on Mardia's circular information inequality. The proposed cyclic CRBs are lower bounds on the MCE of any cyclic-unbiased estimator. One of these bounds has been recently developed by using the cyclic Hammersley-Chapman-Robbins lower bounds on the MCE. The derivations presented in this paper relates between the cyclic CRB and existing results from directional statistics on manifolds. The properties of the proposed cyclic CRBs are examined. In particular, it is shown that the cyclic CRBs are always lower than the convectional CRB and that the two cyclic CRBs are asymptotically achievable by the performance of the maximum likelihood (ML) estimator. The proposed cyclic CRBs and the performance of the ML estimator are compared in terms of MCE in estimation of the mean of von Mises distributed measurements and for phase estimation with additive white Gaussian noise (AWGN).

[1]  João M. F. Xavier,et al.  Intrinsic distance lower bound for unbiased estimators on Riemannian manifolds , 2002, 2002 IEEE International Conference on Acoustics, Speech, and Signal Processing.

[2]  Ruggero Reggiannini,et al.  A fundamental lower bound to the performance of phase estimators over Rician-fading channels , 1997, IEEE Trans. Commun..

[3]  H. Cramér.  A contribution to the theory of statistical estimation , 1946 .

[4]  H. Hendriks.  A Crame´r-Rao–type lower bound for estimators with values in a manifold , 1991 .

[5]  Yoram Bresler,et al.  A global lower bound on parameter estimation error with periodic distortion functions , 2000, IEEE Trans. Inf. Theory.

[6]  Brian C. Lovell,et al.  The circular nature of discrete-time frequency estimates , 1991, [Proceedings] ICASSP 91: 1991 International Conference on Acoustics, Speech, and Signal Processing.

[7]  Gerhard Kurz,et al.  Trigonometric moment matching and minimization of the Kullback-Leibler divergence , 2015, IEEE Transactions on Aerospace and Electronic Systems.

[8]  B.C. Lovell,et al.  The statistical performance of some instantaneous frequency estimators , 1992, IEEE Trans. Signal Process..

[9]  M. Kupperman.  Linear Statistical Inference and Its Applications 2nd Edition (C. Radhakrishna Rao) , 1975 .

[10]  Arye Nehorai,et al.  Vector-sensor array processing for electromagnetic source localization , 1994, IEEE Trans. Signal Process..

[11]  Joseph Tabrikian,et al.  MIMO-AR System Identification and Blind Source Separation for GMM-Distributed Sources , 2009, IEEE Transactions on Signal Processing.

[12]  R. S. Bucy,et al.  An Optimal Phase Demodulator , 1975 .

[13]  Uwe D. Hanebeck,et al.  Efficient Nonlinear Bayesian Estimation based on Fourier Densities , 2006, 2006 IEEE International Conference on Multisensor Fusion and Integration for Intelligent Systems.

[14]  K. Mardia.  Statistics of Directional Data , 1972 .

[15]  Arye Nehorai,et al.  Performance bounds for estimating vector systems , 2000, IEEE Trans. Signal Process..

[16]  Joseph Tabrikian,et al.  Bayesian Parameter Estimation Using Periodic Cost Functions , 2012, IEEE Transactions on Signal Processing.

[17]  Joseph Tabrikian,et al.  Mean-cyclic-error lower bounds via integral transform of likelihood-ratio function , 2016, 2016 IEEE Sensor Array and Multichannel Signal Processing Workshop (SAM).

[18]  A. Neath.  Testing Statistical Hypotheses (3rd ed.). E. L. Lehmann and Joseph P. Romano , 2006 .

[19]  Ioannis Pitas,et al.  Nonlinear processing and analysis of angular signals , 1998, IEEE Trans. Signal Process..

[20]  James Ting-Ho Lo,et al.  Exponential Fourier densities and optimal estimation and detection on the circle , 1977, IEEE Trans. Inf. Theory.

[21]  H. V. Trees,et al.  Covariance, Subspace, and Intrinsic CramrRao Bounds , 2007 .

[22]  F. Ruymgaart,et al.  A Cramér-Rao type inequality for random variables in Euclidean manifolds , 1992 .

[23]  Andrew R. Barron,et al.  Mixture Density Estimation , 1999, NIPS.

[24]  I. Vaughan L. Clarkson,et al.  Direction Estimation by Minimum Squared Arc Length , 2012, IEEE Transactions on Signal Processing.

[25]  Joseph Tabrikian,et al.  General Classes of Performance Lower Bounds for Parameter Estimation—Part I: Non-Bayesian Bounds for Unbiased Estimators , 2010, IEEE Transactions on Information Theory.

[26]  Dave Zachariah,et al.  Line spectrum estimation with probabilistic priors , 2013, Signal Process..

[27]  Robert Boorstyn,et al.  Single tone parameter estimation from discrete-time observations , 1974, IEEE Trans. Inf. Theory.

[28]  P. Willett,et al.  CRLB for likelihood functions with parameter-dependent support and a new bound , 2014, 2014 IEEE Aerospace Conference.

[29]  Joseph Tabrikian,et al.  Bayesian cyclic bounds for periodic parameter estimation , 2013, 2013 5th IEEE International Workshop on Computational Advances in Multi-Sensor Adaptive Processing (CAMSAP).

[30]  Joseph Tabrikian,et al.  Cyclic Barankin-Type Bounds for Non-Bayesian Periodic Parameter Estimation , 2014, IEEE Transactions on Signal Processing.

[31]  Steven Kay,et al.  Unbiased estimation of the phase of a sinusoid , 2004, 2004 IEEE International Conference on Acoustics, Speech, and Signal Processing.

[32]  Joseph Tabrikian,et al.  Non-Bayesian Periodic Cramér-Rao Bound , 2013, IEEE Transactions on Signal Processing.

[33]  Joseph Tabrikian,et al.  A New Class of Bayesian Cyclic Bounds for Periodic Parameter Estimation , 2016, IEEE Transactions on Signal Processing.

[34]  Joseph Tabrikian,et al.  Periodic CRB for non-Bayesian parameter estimation , 2011, 2011 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP).

[35]  S.T. Smith,et al.  Covariance, subspace, and intrinsic Crame/spl acute/r-Rao bounds , 2005, IEEE Transactions on Signal Processing.

[36]  Kanti V. Mardia,et al.  Statistics of Directional Data , 1972 .

[37]  Yonina C. Eldar.  MSE Bounds With Affine Bias Dominating the CramÉr–Rao Bound , 2008, IEEE Transactions on Signal Processing.

[38]  Stephen E. Fienberg,et al.  Testing Statistical Hypotheses , 2005 .

[39]  Eric Chaumette,et al.  A New Barankin Bound Approximation for the Prediction of the Threshold Region Performance of Maximum Likelihood Estimators , 2008, IEEE Transactions on Signal Processing.

[40]  Philipp Berens,et al.  CircStat: AMATLABToolbox for Circular Statistics , 2009, Journal of Statistical Software.

[41]  E. Barankin.  Locally Best Unbiased Estimates , 1949 .

[42]  Joseph Tabrikian,et al.  On the limitations of Barankin type bounds for MLE threshold prediction , 2015, Signal Process..

[43]  Calyampudi Radhakrishna Rao,et al.  Linear Statistical Inference and its Applications , 1967 .

[44]  Marvin K. Simon,et al.  Digital Communication Techniques: Signal Design and Detection , 2008 .

[45]  Eric Chaumette,et al.  New Results on Deterministic Cramér–Rao Bounds for Real and Complex Parameters , 2012, IEEE Transactions on Signal Processing.

[46]  Joseph Tabrikian,et al.  Cyclic Bayesian Cramér-Rao bound for filtering in circular state space , 2015, 2015 18th International Conference on Information Fusion (Fusion).

[47]  Philippe Forster,et al.  On lower bounds for deterministic parameter estimation , 2002, 2002 IEEE International Conference on Acoustics, Speech, and Signal Processing.

[48]  T. Hotz,et al.  Intrinsic means on the circle: uniqueness, locus and asymptotics , 2011, 1108.2141.

[49]  I. Vaughan L. Clarkson,et al.  Analysis of the variance threshold of Kay's weighted linear predictor frequency estimator , 1994, IEEE Trans. Signal Process..