Title :
Time-varying autoregressive modeling of HRR radar signatures
Author_Institution :
Dept. of Electr. Eng. & Comput. Sci., George Washington Univ., Washington, DC, USA
fDate :
7/1/1999 12:00:00 AM
Abstract :
A time-varying autoregressive (TVAR) model is used for the modeling and classification of high range resolution (HRR) radar signatures. In this approach, the TVAR coefficients are expanded by a low-order discrete Fourier transform (DFT). A least-squares (LS) estimator of the TVAR model parameters is presented, and the maximum likelihood (ML) approach for determining the model order is also presented. The validity of the TVAR modeling approach is demonstrated by comparing with other approaches in estimating time-varying spectra of synthetic signals. The estimated TVAR model parameters are also used as features in classifying HRR radar signatures with a neural network. In the experiment with two sets of noncooperating target identification (NCTI) data, about 93% of samples are correctly classified
Keywords :
autoregressive processes; discrete Fourier transforms; feature extraction; least squares approximations; maximum likelihood estimation; modelling; neural nets; radar resolution; radar target recognition; signal classification; signal representation; time series; time-frequency analysis; PCA; automatic target recognition; feature extraction; feature space reduction; high range resolution radar signatures; least-squares estimator; low-order discrete Fourier transform; maximum likelihood approach; model order; neural network classification; noncooperating target identification data; parameter estimation; radar signatures classification; time-frequency distribution; time-varying autoregressive model; Application software; Autoregressive processes; Discrete Fourier transforms; Feature extraction; Maximum likelihood estimation; Millimeter wave radar; Polynomials; Radar applications; Radar clutter; Radar signal processing;
Journal_Title :
Aerospace and Electronic Systems, IEEE Transactions on
