Performance Analysis of a Parameterized APES (PAPES) Spectrum Estimation Method for ISAR Applications

摘要

Spectral analysis algorithms are divided in two major categories: parametric and non-parametric, where the latter apply a narrow band-pass filter to the input signal and the filter output power divided by the filter bandwidth gives a measure of the input signal's spectrum [1]. Many of these methods, like the Capon [2] and the APES [3] estimators, belong to the matched-filterbank (MAFI) class [4], which achieve spectral estimation by using adaptive finite impulse response (FIR) filterbanks. Also the C&A algorithm for the estimation of the spectrum of spectral lines was introduced in [5]. Complex spectral estimation methods are important to the Inverse Synthetic Aperture Radar (ISAR) imaging technique which produces 2-D high resolution images of moving targets [6-7]. ISAR research is also related with the development of image reconstruction [8-9], motion compensation and image focusing techniques [10-11]. The use of CAPON and APES for ISAR imaging is discussed in [2], [3] and [12].