ResearchTrend.AI
  • Papers
  • Communities
  • Events
  • Blog
  • Pricing
Papers
Communities
Social Events
Terms and Conditions
Pricing
Parameter LabParameter LabTwitterGitHubLinkedInBlueskyYoutube

© 2025 ResearchTrend.AI, All rights reserved.

  1. Home
  2. Papers
  3. 2010.04345
25
12
v1v2 (latest)

Exact Minimax Estimation for Phase Synchronization

9 October 2020
Chao Gao
A. Zhang
ArXiv (abs)PDFHTML
Abstract

We study the phase synchronization problem with measurements Y=z∗z∗H+σW∈Cn×nY=z^*z^{*H}+\sigma W\in\mathbb{C}^{n\times n}Y=z∗z∗H+σW∈Cn×n, where z∗z^*z∗ is an nnn-dimensional complex unit-modulus vector and WWW is a complex-valued Gaussian random matrix. It is assumed that each entry YjkY_{jk}Yjk​ is observed with probability ppp. We prove that the minimax lower bound of estimating z∗z^*z∗ under the squared ℓ2\ell_2ℓ2​ loss is (1−o(1))σ22p(1-o(1))\frac{\sigma^2}{2p}(1−o(1))2pσ2​. We also show that both generalized power method and maximum likelihood estimator achieve the error bound (1+o(1))σ22p(1+o(1))\frac{\sigma^2}{2p}(1+o(1))2pσ2​. Thus, σ22p\frac{\sigma^2}{2p}2pσ2​ is the exact asymptotic minimax error of the problem. Our upper bound analysis involves a precise characterization of the statistical property of the power iteration. The lower bound is derived through an application of van Trees' inequality.

View on arXiv
Comments on this paper