MUSIC for multidimensional spectral estimation: stability and super-resolution

TL;DR

This paper analyzes the stability and super-resolution capabilities of the MUSIC algorithm for multidimensional spectral estimation from single snapshots, providing theoretical guarantees and noise robustness estimates.

Contribution

It offers a comprehensive performance analysis of MUSIC in multidimensional settings, including stability bounds and super-resolution limits under noise.

Findings

Exact reconstruction with $(2s)^D$ measurements in noiseless case.

Explicit noise perturbation estimates based on noise level and frequency separation.

Noise tolerance decays as $rac{ ext{constant}}{ ext{sample size}^{1/2}}$ under Gaussian noise.

Abstract

This paper presents a performance analysis of the MUltiple SIgnal Classification (MUSIC) algorithm applied on $D$ dimensional single-snapshot spectral estimation while $s$ true frequencies are located on the continuum of a bounded domain. Inspired by the matrix pencil form, we construct a D-fold Hankel matrix from the measurements and exploit its Vandermonde decomposition in the noiseless case. MUSIC amounts to identifying a noise subspace, evaluating a noise-space correlation function, and localizing frequencies by searching the $s$ smallest local minima of the noise-space correlation function. In the noiseless case, $(2s)^D$ measurements guarantee an exact reconstruction by MUSIC as the noise-space correlation function vanishes exactly at true frequencies. When noise exists, we provide an explicit estimate on the perturbation of the noise-space correlation function in terms of noise level, dimension $D$, the minimum separation among frequencies, the maximum and minimum amplitudes while frequencies are separated by two Rayleigh Length (RL) at each direction. As a by-product the maximum and minimum non-zero singular values of the multidimensional Vandermonde matrix whose nodes are on the unit sphere are estimated under a gap condition of the nodes. Under the 2-RL separation condition, if noise is i.i.d. gaussian, we show that perturbation of the noise-space correlation function decays like $\sqrt{\log(\#(\mathbf{N}))/\#(\mathbf{N})}$ as the sample size $\#(\mathbf{N})$ increases. When the separation among frequencies drops below 2 RL, our numerical experiments show that the noise tolerance of MUSIC obeys a power law with the minimum separation of frequencies.