Gridless variational Bayesian inference of line spectral from quantized samples

Efficient estimation of line spectral from quantized samples is significant importance in information theory and signal processing, e.g., channel energy efficient massive MIMO systems direction arrival estimation. The goal this paper to recover the as well its corresponding parameters including model order, frequencies amplitudes heavily samples. To end, we propose an grid-less Bayesian algorithm named VALSE-EP, which a combination high resolution low complexity gridless variational (VALSE) expectation propagation (EP). basic idea VALSE-EP iteratively approximate challenging sequence simple pseudo unquantized models, where VALSE applied. Moreover, obtain benchmark performance proposed algorithm, Cramer Rao bound (CRB) derived. Finally, numerical experiments on both synthetic real data are performed, demonstrating near CRB for