2-D data extrapolation for high resolution radar imaging using autoregressive lattice modelling

A now method to estimate backscattered fields beyond the measurement range is described. The available data is extrapolated by 2-D linear prediction o f 2-D cartesian frequency spectra using 2-D orthogonal lattice predictors. Since this technique does not guarantee a stable prediction filter, one may have to modify the prediction parameters to ensure stability. These modifi ed parameters can be used for the 2-D extrapolation of the 2-D cartesian ba ckscattered data. It is shown that the inverse Fourier transform of the ext ended data achieves better resolved images both in range and cross-range fo r simulated and experimental targets. The algorithm is based on 2-D extrapo lation of the backscattered data, while other existing algorithms extrapola te range or cross-range profiles separately owing to the separability assum ption of the input data. The method provides more reliable results for both experimental and real-world targets which cannot be described by the point target model. Moreover, the complexity related with the calculation of pre diction coefficients is independent of the extrapolation factor.