Variational Bayesian Inference of Line Spectral Estimation with Multiple Measurement Vectors

In this paper, the line spectral estimation (LSE) problem with multiple measurement vectors (MMVs) is studied utilizing the Bayesian methods. Motivated by the recently proposed variational line spectral estimation (VALSE) method, we extend it to deal with the MMVs setting, which is especially important in array signal processing. The VALSE method can automatically estimate the model order and nuisance parameters such as noise variance and weight variance. In addition, by approximating the probability density function (PDF) of the frequencies with the mixture of von Mises PDFs, closed-form update equation and the uncertainty degree of the estimates can be obtained. Interestingly, we find that the VALSE with MMVs can be viewed as applying the VALSE with single measurement vector (SMV) to each snapshot, and combining the intermediate data appropriately. Furthermore, the proposed prior distribution provides a good interpretation of tradeoff between grid and off-grid based methods. Finally, numerical results demonstrate the effectiveness of the VALSE method, compared to the state-of-the-art methods in the MMVs setting.