Imaging of a compact range using autoregressive spectral estimation

An estimation technique based on the autoregressive (AR) modeling of field probe data is used to locate and quantify spurious signals in a compact range. In this technique, the probe aperture is divided into a number of overlapping subapertures such that the far-field criterion for each subaperture is satisfied. Then the subaperture data are modeled as an AR process, and the AR parameters are derived using the principal component forward-backward linear prediction technique. Directions of the incident signals relative to each subaperture are then determined from the poles of the prediction filters. Using a series of subapertures, the locations of the scatterers are estimated by triangulation. After estimation of the spatial frequencies of the probe data for any subaperture, the magnitude of each component is determined by a least squares algorithm. Examples of probe measurements and analysis for the Ohio State University compact range are given