Abstract
A new method for permittivity profile reconstruction based on terahertz time-domain spectroscopy signal processing is presented. Reconstruction is accomplished in two steps. First, the sample pulse function is reconstructed using sample time-domain reflection data. Low and high frequency noise filtering and the interpolation of the pulse function at low frequencies are then applied. Second, an invariant embedding technique is used to calculate the dielectric permittivity profile based on the sample pulse function. Method was tested experimentally on samples with known permittivity profiles. The algorithm is stable to additive Gaussian white noise as shown using mathematical modeling based on the finite-difference time-domain technique (FDTD). Possible applications of this permittivity profile reconstruction technique are discussed.
Export citation and abstract BibTeX RIS
Content from this work may be used under the terms of the Creative Commons Attribution 3.0 licence. Any further distribution of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI.