Analysis of Fisher Information and the Cramér–Rao Bound for Nonlinear Parameter Estimation After Random Compression

In this paper, we analyze the impact of compression with complex random matrices on Fisher information and the Cramér-Rao Bound (CRB) for estimating unknown parameters in the mean value function of a complex multivariate normal distribution. We consider the class of random compression matrices whose distribution is right-unitarily invariant. The compression matrix whose elements are i.i.d. standard complex normal random variables is one such matrix. We show that for all such compression matrices, the Fisher information matrix has a complex matrix beta distribution. We also derive the distribution of CRB. These distributions can be used to quantify the loss in CRB as a function of the Fisher information of the noncompressed data. In our numerical examples, we consider a direction of arrival estimation problem and discuss the use of these distributions as guidelines for choosing compression ratios based on the resulting loss in CRB.

[1]  Massimo Fornasier,et al.  Compressive Sensing , 2015, Handbook of Mathematical Methods in Imaging.

[2]  L. Scharf,et al.  Geometry of the Cramer-Rao bound , 1992, [1992] IEEE Sixth SP Workshop on Statistical Signal and Array Processing.

[3]  S. Frick,et al.  Compressed Sensing , 2014, Computer Vision, A Reference Guide.

[4]  Edwin K. P. Chong,et al.  Compressive Sensing and Sparse Inversion in Signal Processing: Cautionary Notes , 2011 .

[5]  E.J. Candes.  Compressive Sampling , 2022 .

[6]  Upamanyu Madhow,et al.  Compressive estimation in AWGN: General observations and a case study , 2012, 2012 Conference Record of the Forty Sixth Asilomar Conference on Signals, Systems and Computers (ASILOMAR).

[7]  Louis L. Scharf,et al.  Geometry of the Cramer-Rao bound , 1993, Signal Process..

[8]  R.G. Baraniuk,et al.  Compressive Sensing [Lecture Notes] , 2007, IEEE Signal Processing Magazine.

[9]  A. James.  Distributions of Matrix Variates and Latent Roots Derived from Normal Samples , 1964 .

[10]  Edwin K. P. Chong,et al.  Sensitivity considerations in compressed sensing , 2011, 2011 Conference Record of the Forty Fifth Asilomar Conference on Signals, Systems and Computers (ASILOMAR).

[11]  Steven Kay,et al.  Fundamentals Of Statistical Signal Processing , 2001 .

[12]  Muni S. Srivastava.  On the Complex Wishart Distribution , 1965 .

[13]  Yuejie Chi,et al.  Analysis of fisher information and the Cramer-Rao bound for nonlinear parameter estimation after compressed sensing , 2013, 2013 IEEE International Conference on Acoustics, Speech and Signal Processing.

[14]  Y. Chikuse.  Statistics on special manifolds , 2003 .

[15]  A. Robert Calderbank,et al.  Sensitivity to Basis Mismatch in Compressed Sensing , 2011, IEEE Trans. Signal Process..

[16]  Søren Holdt Jensen,et al.  On compressed sensing and the estimation of continuous parameters from noisy observations , 2012, 2012 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP).

[17]  Christian Jutten,et al.  On the Achievability of Cramér–Rao Bound in Noisy Compressed Sensing , 2012, IEEE Transactions on Signal Processing.

[18]  K Gupta Arjun,et al.  PROPERTIES OF THE COMPLEX MATRIX VARIATE DIRICHLET DISTRIBUTION , 2007 .

[19]  Vahid Tarokh,et al.  Asymptotic Achievability of the CramÉr–Rao Bound for Noisy Compressive Sampling , 2009, IEEE Transactions on Signal Processing.

[20]  I. Reed,et al.  Rapid Convergence Rate in Adaptive Arrays , 1974, IEEE Transactions on Aerospace and Electronic Systems.