Grid-less variational Bayesian line spectral estimation with multiple measurement vectors

Line spectral estimation (LSE) with multiple measurement vector (MMV) is studied utilizing the Bayesian variational inference. Motivated by the recent grid-less variational line spectral estimation (VALSE) method, we develop the MMV VALSE (MVALSE). The MVALSE shares the advantages of the VALSE method, such as automatically estimating the model order, noise variance, weight variance, and providing the uncertainty of the frequency estimates. The MVALSE can be viewed as applying the VALSE with single measurement vector to each snapshot, and combining the intermediate data appropriately. Furthermore, the MVALSE is developed to perform sequential estimation. Numerical results demonstrate the effectiveness of the MVALSE method, compared to the state-of-the-art MMV methods.