During the diving movement for highly squinted synthetic aperture radar (SAR)

the echo is coupled greatly between azimuth and range

invariance in azimuth is not accurate

which are caused by high vertical velocity and acceleration

making SAR imaging difficult to process.Due to these issues

a frequency phase filtering algorithm (FPFA) is proposed.First

the accurate 2-D spectrum can be got via the method of series reversion (MSR)

the range walk will be eliminated in the time domain and in the frequency domain the range curvature removed.Then

a frequency filtering phase is leaded into frequency domain in order to correct the dependence of the FM rates.At last

the signal is focused in the Doppler domain by Spectral Analysis technique (SPECAN).As the whole procedure only involves short fast Fourier transforms (FFTs and complex multiplications without interpolation

the FPFA is computationally efficient and facilitates hardware real-time implementation.The simulated results and raw data processing validate the effectiveness of the proposed algorithm.