Almost sure well-posedness and scattering of the 3D cubic nonlinear Schrödinger equation

Abstract: We study the random data problem for 3D, defocusing, cubic nonlinear Schr\"odinger equation in $H_x^s(\mathbb{R}^3)$ with $s<\frac 12$. First, we prove that the almost sure local well-posedness holds when $\frac{1}{6}\leqslant s<\frac 12$ in the sense that the Duhamel term belongs to $H_x^{1/2}(\mathbb{R}^3)$. Furthermore, we prove that the global well-posedness and scattering hold for randomized, radial, large data $f\in H_x^{s}(\mathbb{R}^3)$ when $\frac{17}{40}< s<\frac 12$. The key ingredient is to control the energy increment including the terms where the first order derivative acts on the linear flow, and our argument can lower down the order of derivative more than $\frac12$. To our best knowledge, this is the first almost sure large data global result for this model.