Identification of AR systems at a very low SNR using damped cosine model of autocorrelation function

S Anowarul Fattah,M Rezwan Khan,Md Kamrul Hasan

doi:10.1109/icassp.2002.5744957

Identification of AR systems at a very low SNR using damped cosine model of autocorrelation function

S Anowarul Fattah, M Rezwan Khan + Show 1 more

https://doi.org/10.1109/icassp.2002.5744957

Copy DOI

Publication Date: May 1, 2002

Citations: 5

Affiliation: Bangladesh University of Engineering and Technology

#Signal To Noise Ratio #Identification Of Autoregressive Systems + Show 8 more

Abstract
Full-Text PDF
Similar Papers

Abstract

This paper presents a new method for autoregressive (AR) system identification at a very low signal to noise ratio (SNR) using damped cosine model for the autocorrelation function of the noise-free, AR signal. The AR parameters are obtained directly from the estimated damped cosine model parameters. The simulation results show that the method can estimate the system parameters with high accuracy even at an SNR as low as −5 dB.

Full Text