Table of Contents
Fetching ...

Sums of projections with random coefficients

Leonid Pastur, Alexander Pushnitski

TL;DR

This work analyzes infinite sums of random rank-one projections ${\mathcal P}_{\varkappa}=\sum_{n} \varkappa_n \langle\cdot,\psi_n\rangle\psi_n$ in a Hilbert space, with a translation-invariant Gram matrix $G$ encoded by the symbol $\varphi$ through the Laurent operator $L_{\varphi}$. By exploiting ergodicity, they prove the spectrum is almost surely deterministic and define the integrated density of states $\nu_{\mathcal{P}_{\varkappa}}$, establishing a Wegner-type bound and Lifshitz tails for the IDS; they also derive explicit moment formulas linking $\nu_{\mathcal{P}_{\varkappa}}$ to $\mathbf E[\varkappa_0]$, $\mathbf V[\varkappa_0]$, and the Fourier coefficients of $\varphi$. The paper provides concrete realizations in $\ell^2(\mathbb{Z})$ and $L^2(\mathbb{T})$, relates the spectrum to multiplicative perturbations of Laurent operators, and proves Anderson localisation in the nearly-orthogonal regime via fractional moments and Simon–Wolff techniques. These results extend ergodic operator theory to a nonlocal random-projection setting and connect the spectral properties to a rich structure of the Gram symbol and Beurling–Mallivin theory for possible atoms in the IDS. Overall, the work offers a rigorous framework for understanding randomness in sums of projections, with implications for nonlocal random operators and ergodic spectral theory.

Abstract

We study infinite sums \[ {\mathcal P}_{\varkappa}=\sum_{n=-\infty}^\infty \varkappa_n \langle\cdot, ψ_n\rangleψ_n \] of rank-one projections in a Hilbert space, where $\{ψ_n\}_{n\in\mathbb Z}$ are norm-one vectors, not necessarily orthogonal, and $\{\varkappa_n\}_{n\in\mathbb Z}$ are independent identically distributed positive random variables. Assuming that the Gram matrix $\{\langleψ_n,ψ_m\rangle\}_{n,m\in\mathbb Z}$ defines a bounded operator on $\ell^2(\mathbb Z)$ and that its entries depend only on the difference $n-m$, we analyse ${\mathcal P}_{\varkappa}$ within the framework of spectral theory of ergodic operators. Inspired by the spectral theory of ergodic Schrödinger operators, we define the integrated density of states (IDS) measure $ν_{{\mathcal P}_\varkappa}$ for ${\mathcal P}_{\varkappa}$ and establish results on its continuity and absolute continuity, including Wegner-type estimates and Lifshitz tail behaviour near the spectral edges. In the asymptotic regime of nearly-orthogonal $ψ_n$, we prove the Anderson-type localisation result: the spectrum of ${\mathcal P}_{\varkappa}$ is pure point almost surely.

Sums of projections with random coefficients

TL;DR

This work analyzes infinite sums of random rank-one projections in a Hilbert space, with a translation-invariant Gram matrix encoded by the symbol through the Laurent operator . By exploiting ergodicity, they prove the spectrum is almost surely deterministic and define the integrated density of states , establishing a Wegner-type bound and Lifshitz tails for the IDS; they also derive explicit moment formulas linking to , , and the Fourier coefficients of . The paper provides concrete realizations in and , relates the spectrum to multiplicative perturbations of Laurent operators, and proves Anderson localisation in the nearly-orthogonal regime via fractional moments and Simon–Wolff techniques. These results extend ergodic operator theory to a nonlocal random-projection setting and connect the spectral properties to a rich structure of the Gram symbol and Beurling–Mallivin theory for possible atoms in the IDS. Overall, the work offers a rigorous framework for understanding randomness in sums of projections, with implications for nonlocal random operators and ergodic spectral theory.

Abstract

We study infinite sums of rank-one projections in a Hilbert space, where are norm-one vectors, not necessarily orthogonal, and are independent identically distributed positive random variables. Assuming that the Gram matrix defines a bounded operator on and that its entries depend only on the difference , we analyse within the framework of spectral theory of ergodic operators. Inspired by the spectral theory of ergodic Schrödinger operators, we define the integrated density of states (IDS) measure for and establish results on its continuity and absolute continuity, including Wegner-type estimates and Lifshitz tail behaviour near the spectral edges. In the asymptotic regime of nearly-orthogonal , we prove the Anderson-type localisation result: the spectrum of is pure point almost surely.

Paper Structure

This paper contains 50 sections, 219 equations.

Theorems & Definitions (13)

  • proof
  • proof
  • proof : Proof of Theorem \ref{['thm:aa4']}
  • proof
  • proof
  • proof
  • proof
  • proof
  • proof
  • proof
  • ...and 3 more