Estimation of noisy quantized random observation coefficient AR time-series

Author

Krishnamurthy, Vikram ; Mareels, I.M.Y.

Author_Institution

Dept. of Syst. Eng., Australian Nat. Univ., Canberra, ACT, Australia

fYear

1994

fDate

27 Jun-1 Jul 1994

Firstpage

115

Abstract

We present a consistent, asymptotically normal estimation algorithm for the parameters of auto-regressive (AR) processes from 1-bit quantized observations. The input signal to the quantifier is the AR signal corrupted by multiplicative white Gaussian noise. Our algorithm is computationally inexpensive as it involves counting the number of occurrences of particular patterns of zeros and ones in the observation sequence and then solving a Yule-Walker type system

Keywords

Gaussian noise; autoregressive processes; parameter estimation; quantisation (signal); sequential estimation; time series; white noise; 1-bit quantized observations; AR signal; AR time-series; Yule-Walker type system; asymptotically normal estimation algorithm; auto-regressive processes; input signal; multiplicative white Gaussian noise; noisy quantized random observation coefficient; observation sequence; parameter estimation; Australia; Delta modulation; Equations; Gaussian noise; Signal processing; Speech analysis; Speech coding; Systems engineering and theory; Time series analysis; Yttrium;

fLanguage

English

Publisher

ieee

Conference_Titel

Information Theory, 1994. Proceedings., 1994 IEEE International Symposium on

Conference_Location

Trondheim

Print_ISBN

0-7803-2015-8

Type

conf

DOI

10.1109/ISIT.1994.394873

Filename

394873