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

A new method to estimate backscattered fields beyond the measurement range is described. The available data is extrapolated by 2-D linear prediction of 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 modified parameters can be used for the 2-D extrapolation of the 2-D cartesian backscattered data. It is shown that the inverse Fourier transform of the extended data achieves better resolved images both in range and cross-range for simulated and experimental targets. The algorithm is based on 2-D extrapolation of the backscattered data, while other existing algorithms extrapolate range or cross-range profiles separately owing to the separability assumption 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 prediction coefficients is independent of the extrapolation factor.