Local well-posedness for quadratic nonlinear Schrödinger equations and the ``good'' Boussinesq equation. (Q636906): Difference between revisions

The authors consider a class of the one-dimensional quadratic nonlinear Schrödinger equations of the form \((i\partial _t-\partial ^2_x)u=N(u)\) where \(N(u)=\alpha _1 u^2 + \alpha _2\overline {u}^2 + \alpha _3 u\overline {u}\), \(\alpha _1\), \(\alpha _2\), \(\alpha _3\) being complex constants. The authors work in the space \(H^{s,a}\) defined with the help of the Fourier transform norm as the space of all tempered distributions for which the \(L^2\) norm of \((1+| \xi | )^{s-a} | \xi | ^{a} \widehat {f}(\xi )\) is finite. For a Cauchy problem, the authors prove the local well-posedness of the problem in the space \(H^{s,a}\) for \(| a| <\frac {1}{2}\), \(s\geq -\frac {2a+1}{4}\). For \(a=0\) this yields a local well-posedness in \(H^{s}\) for \(s\geq -\frac {1}{4}\), a new result. As a consequence, the authors obtain a local well-posedness result for the ``good'' Boussinesq equation in \(H^{s}\times H^{s-2}\) with \(s>-\frac {1}{2}\), an improvement of the previous result by \textit{L. G. Farah} [Commun. Partial Differ. Equations 34, No. 1, 52--73 (2009; Zbl 1173.35682)]. The paper is clearly written, even if it is rather technical.