Gabor Frame Decomposition of Evolution Operators and Applications

Michele Berra

DOI: https://doi.org/10.1007/s11868-014-0090-8

2014-02-26

Abstract:We compute the Gabor matrix for Schrödinger-type evolution operators. Precisely, we analyze the Heat Equation, already presented in \cite{2012arXiv<a class="link-https" data-arxiv-id="1209.0945" href="https://arxiv.org/abs/1209.0945">1209.0945</a>C}, giving the exact expression of the Gabor matrix which leads to better numerical evaluations. Then, using asymptotic integration techniques, we obtain an upper bound for the Gabor matrix in one-dimension for the generalized Heat Equation, new in the literature. Using Maple software, we show numeric representations of the coefficients' decay. Finally, we show the super-exponential decay of the coefficients of the Gabor matrix for the Harmonic Repulsor, together with some numerical evaluations. This work is the natural prosecution of the ideas presented in \cite{2012arXiv<a class="link-https" data-arxiv-id="1209.0945" href="https://arxiv.org/abs/1209.0945">1209.0945</a>C} and \cite{MR2502369}.

Analysis of PDEs

What problem does this paper attempt to address?

The main problem that this paper attempts to solve is to study the time - frequency representations of Schrödinger - type evolution operators by computing their Gabor matrices and to provide numerical evidence for the decay properties of the coefficients. Specifically:

1. **Heat Equation**: The author improved the Gabor matrix estimates for the heat equation in previous work and gave an exact expression, thereby increasing the decay rate of the coefficients.

2. **Generalized Heat Equation**: For the first time in the literature, the author gave an upper - bound estimate of the Gabor matrix of the generalized heat equation and analyzed in detail the constants involved, providing more accurate decay information.

3. **Harmonic Repeller**: The author analyzed the Gabor matrix of the harmonic repeller, gave its exact representation, and proved the super - exponential decay property of the coefficients, similar to the case of the harmonic oscillator.

Through these studies, the author aims to provide an effective method to accelerate the calculation of the action of evolution operators, especially when dealing with large - scale data, by selecting important matrix coefficients to reduce the computational complexity. The paper also shows numerical simulations using Maple software to verify the theoretical results.