A Novel Airborne Radar Forward-looking Imaging Algorithm Based on Deconvolution

The performance of forward-looking imaging for airborne radar is mostly determined by the azimuth resolution, which can be improved by deconvolution algorithm. However, because of the ill-posed problem caused by direct deconvolution processing in azimuth dimension, the enhancement of azimuth resolution is actually limited so far. In this paper, a linear prediction method based on the optimal criterion of minimum mean square prediction error is proposed to solve this problem and accurately reconstruct the ground scattering coefficients. As is validated by simulation and real data processing results, the deconvolution algorithm based on linear prediction (LP) has a remarkable improvement on azimuth resolution, leading to a better imaging performance as compared to the original method.