A Super-Resolution Parameter Estimation Algorithm for Multi- Dimensional NMR Spectroscopy

Abstract

In this paper, we will propose a super-resolution scheme for the parameter estimation of multi-dimensional (M-D) NMR spectroscopy. M-D NMR signals can be modeled as the summation of M-D damped sinusoids. The frequencies and the damping factors of M-D damped sinusoids play important roles in protein structure determination using M-D NMR spectroscopy. We will develop a super-resolution frequency and damping factor estimation algorithm-damped MUSIC (DMUSIC) algorithm. Since the DMUSIC algorithm makes full use of the rank-deficiency and the Hankel property of the data matrix composed of the M-D NMR data, compared with other NMR data analysis algorithms, it can resolve the spectrum using very few data points. The performance of the DMUSIC algorithm is demonstrated by computer simulations.