Modeling of non-gaussian AR model with transient coefficients using wavelet basis

Authors:
Lisha Sun;Minfen Shen;Weiling Xu;Patch Beadle
Affiliations:
Dept. of Electronic Engineering, Shantou University, Guangdong, China;Dept. of Electronic Engineering, Shantou University, Guangdong, China;Dept. of Electronic Engineering, Shantou University, Guangdong, China;School of System Engineering, Portsmouth University, Portsmouth, U.K.
Venue:
PDCAT'04 Proceedings of the 5th international conference on Parallel and Distributed Computing: applications and Technologies
Year:
2004

Citing 3
Cited 0

Results on AR-modelling of nonstationary signals

Signal Processing
Ten lectures on wavelets

Ten lectures on wavelets
New cumulant-based approaches for non-Gaussian time-varying AR models

Signal Processing

Quantified Score

Hi-index	0.00

Visualization

Abstract

This paper focuses on the modeling of non-Gaussian autoregressive (AR) model for a wide range of physical signals. A practical algorithm based on higher-order cumulants is proposed to deal with the problem of estimating the non-Gaussian AR model with transient coefficients. Wavelet basis is used to identify the transient coefficients. The performance in terms of Haar and Morlet basis is evaluated with non-stationary processes. The experimental results show the flexibility of capturing the local events by using the presented model.