Conventional signal processing method

The acquired data can be processed through conventional signal processing tools. Due to the statistical characteristics of the sampled phenomena, random and stationary, a typical approach consists in calculating the Power Spectral Density (PSD) using the Welch method (Welch 1967). 

Solid-state MMW sources with both cryogenically cooled and room temperature detectors are discussed, along with their application to conventional and dielectric perturbation based methods of quantification. Post-detection signal processing, smoothing, filtering and spectral profile fitting are described, which extract the full value from data obtained by this sensitive analytical technique. 

Nestleroth et al. [15], Segal et al. [16] considered some established novel signal processing schemes to assist in adhesive bond inspection. Sinclair et al. [17] and Filimonov [18] employed acoustic resonance methods for dynamic elastic modulus measurements in adhesively bonded structures. Yost and Cantrell [19], Achenbach and Parikh [20] and Nagy et al. [21] considered a nonlinear response of bonded structures to estimate material characteristics. In Achenbach and Parikh [20], failure was preceded by nonlinear behavior of thin boundary layers at the interfaces. Billson and Hutchins [22] considered lasers and EMATS in bond investigations. It was shown that this non-contact technique was reasonable when compared to that obtained by conventional piezoelectric transducers. Ince et al. [23] also characterized bonds with laser-generated ultrasound and through-transmission measurements. 

---