Quantization for spectral super-resolution
Abstract
We show that the method of distributed noise-shaping beta-quantization offers superior performance for the problem of spectral super-resolution with quantization whenever there is redundancy in the number of measurements. More precisely, we define the oversampling ratio as the largest integer such that , where denotes the number of Fourier measurements and is the minimum separation distance associated with the atomic measure to be resolved. We prove that for any number of quantization levels available for the real and imaginary parts of the measurements, our quantization method combined with either TV-min/BLASSO or ESPRIT guarantees reconstruction accuracy of order and respectively, where the implicit constants are independent of , and . In contrast, naive rounding or memoryless scalar quantization for the same alphabet offers a guarantee of order only, regardless of the reconstruction algorithm.
Cite
@article{arxiv.2103.00079,
title = {Quantization for spectral super-resolution},
author = {C. Sinan Güntürk and Weilin Li},
journal= {arXiv preprint arXiv:2103.00079},
year = {2022}
}
Comments
29 pages, 2 figures, to appear in Constructive Approximation