Cable Fault Characterization by Time-Domain Analysis From S-Parameter Measurement and Sparse Inverse Chirp-Z Transform

Cable fault characterization usually requires time-domain profiles to locate and quantify failures. These profiles may be obtained from S-Parameters measured in the frequency domain through inverse Fourier methods. The Chirp Z Transform is a common approach for obtaining time-domain parameters from measured S-parameters. Most existing methods employ the Inverse Chirp Z Transform by a minimum norm solution, which does not have strong physical support. In this paper, we propose a Sparse Inverse Chirp Z Transform approach, specially designed for sparse time-domain signals representing localized fault locations in cables. Two algorithms are proposed to perform the conversion: OMP and FISTA. Their performances are verified in comparison with reference values, showing good agreement for both simulated and experimental data.