Strichartz and dispersive estimates for quantum bouncing ball model: exponential sums and Van der Corput methods in 1d semi-classical Schrödinger equations
Oana Ivanovici
TL;DR
This work analyzes the one-dimensional semi-classical Schrödinger equation on the half-line with a linear potential, focusing on refined dispersive and Strichartz estimates. Central to the approach is a parametrix built from Airy-function spectral data and a reflection-based reformulation that isolates boundary effects, enabling precise control of wave packets. The authors obtain a 1/4-loss dispersive bound in 1D, sharpened via Van der Corput derivative tests to a conjectured $1/6+\epsilon$-loss under sharp exponential-sum bounds, with corresponding improvements in Strichartz estimates. These 1D results illuminate dispersion in strictly convex domains (like the Friedlander model) and offer a foundation for higher-dimensional analyses where tangential dynamics complicate, but the normal-direction losses persist, guiding future work on spectral-cancellation mechanisms and related exponential-sum bounds. The results bridge spectral theory, oscillatory integrals, and number-theoretic sum estimates to advance understanding of quantum bouncing-ball dynamics in semi-classical regimes.
Abstract
We analyze the one-dimensional semi-classical Schrödinger equation on the half-line with a linear potential and Dirichlet boundary conditions. Our main focus is on establishing improved dispersive and Strichartz estimates for this model, which govern the space-time behavior of solutions. We prove refined Strichartz bounds using Van der Corput-type derivative tests, beating previous known results where Strichartz estimates incur 1/4 losses. Moreover, assuming sharp bounds for certain exponential sums, our results indicate the possibility to reduce these losses further to $1/6 + ε$ for all $ε>0$, which would be sharp. We further expect that analogous Strichartz bounds should hold within the Friedlander model domain in higher dimensions.
