Search

A novel terahertz (THz) deconvolution technique based on the autoregressive (AR) spectral extrapolation is presented in this study. An autoregressive process is modeled based on the THz frequency components with high signal-to-noise ratio, and the missing frequency components in regions with low signal-to-noise ratio are extrapolated based on the autoregressive model. In this way, the entire THz frequency spectrum of the impulse response function is estimated. This method is able to provide a ‘quasi-ideal’ impulse response function, and therefore, significantly enhances the depth-resolution for resolving optically thin layers in the THz regime.