Application of Wigner-Ville distribution in electromigration noise analysis

Electromigration noise analysis has shown great potential for the nondestructive evaluation of electromigration performance of metal stripe. However, contradictive conclusions have been published from the electromigration noise analysis. These contradictive conclusions mainly stem from the complex dynamics of the atomic movement during electromigration, rendering the electromigration noise as a nonstationary signal, and, hence, the standard Fourier transform is not adequate. Among the various nonstationary signal analysis tools, Wigner-Ville distribution is used for the analysis of electromigration noise data for the first time. It is found that much "hidden" and useful information in the noise data can be revealed by using this distribution. These information will enable us to "see" the dynamic of the atomic movement during electromigration, enhancing our understanding of electromigration processes.