Nonlinear Correction Approach for FMCW Radar based on Spectrum Recovery

There are two basic approaches for LFM (linear frequency modulation) signal pulse compression – matched filtering and Dechrip. For high-resolution imaging of small scenes, Dechirp is much more popular because it requires a lower sampling rate. However, the transmission signal of demodulation needs to be an ideal chirped signal for Dechirp. When linear modulation cannot be guaranteed accurately, the compression result will be spatially defocused, which causes the compensation complicated. This paper first derives the equivalence of matched filtering and Dechirp processing. Then, for the Dechirp received signal, this paper proposes a pulse compression method for nonlinear FM signals based on matched filtering without upsampling. Finally, the feasibility of the method was verified through experiments.