Phase error autofocus for spotlight SAR imagery

Phase gradient autofocus (PGA) for maximum-likelihood estimation of phase error is most widely used in practice for its efficient realization and excellent robustness. However, because PGA employs adjacent pulses of the range-compressed data to estimate the gradient of phase error, the precision of PGA would be severely affected by the signal-to-noise ratio (SNR) of the range-compressed data. In order to overcome such difficulty, this paper proposes a novel autofocus algorithm to calculate the phase error (PEA) directly. The presented methodology is implemented by weighting average of the input range lines with the fixed phase difference removed. Monte Carlo tests and real SAR data validate that the new approach can achieve better performance than the mostly used PGA with comparable computational cost.