168 / 2019-12-15 14:36:00

Revealing correlations between canonical correlation and correlation coefficient among signal components from multiple datasets

canonical correlation analysis (CCA); canonical correlation coefficients; array signal processing

In this work, we provide a study of canonical correlation analysis (CCA) of data sets from two spatially separated arrays of sensors. Our case studies cover multiple source signals in white noise fields for array signal processing.
The result shows the formula for single source in colored noise also applies to the case of multiple sources in white noise, as long as we factor out the signal of interest component (SOI) when analyzing canonical correlation coefficients. The analytical expression for the canonical correlation coefficients is derived as a function of nominal correlation and signal-to-noise ratio(SNR). Furthermore, we use a direction-of-arrival (DOA) estimation example to show there is a connection between CCA and estimation of signal parameters via rotational invariant techniques (ESPRIT).