Multiplication between Hardy spaces and their dual spaces (Q2334874)

The authors deal with the following problem: For \(p\in (0,1]\) and \(\alpha=1/p-1\), find the ``smallest'' linear vector space \(\mathcal Y\) so that if \(f\in H^p(\mathbb R^n)\) and \(g\in \mathfrak C_\alpha(\mathbb R^n)=(H^p(\mathbb R^n))^\ast\) then \(fg= S(f,g)+ T(f,g)\), where \(S:H^p(\mathbb R^n)\times \mathfrak C_\alpha(\mathbb R^n)\to L^1(\mathbb R^n)\) and \(T:H^p(\mathbb R^n)\times \mathfrak C_\alpha(\mathbb R^n)\to \mathcal Y\) are bounded bilinear maps. This problem for \(p=1\) was initiated by \textit{A. Bonami} et al. [Ann. Inst. Fourier 57, No. 5, 1405--1439 (2007; Zbl 1132.42010)] who found \(\mathcal Y\) to be the weighted Orlicz-Hardy space \(H^\Phi_\omega(\mathbb R^n)\) associated to the weight function \(\omega(x)=1/\log(e+|x|)\) and the Orlicz function \(\Phi(t)=t/\log(e+t)\). \par It was conjectured that the decomposition should be linear in \(f\) and it was solved by \textit{A. Bonami} et al. [J. Math. Pures Appl. (9) 97, No. 3, 230--241 (2012; Zbl 1241.47028)] that actually this was the case when considering \(H^1(\mathbb R^n)\times \mathrm{BMO}(\mathbb R^n)\to L^1(\mathbb R^n)+ H^{\log(\mathbb R^n)}\), where \(H^{\log(\mathbb R^n)}\) stands for the Musielak-Orlicz type space associated to \(\theta(x,t)=t/ (\log(e+t)+\log(e+ |x|))\). In the paper, the case \(0<p<1\) is considered. Concerning linear decompositions they obtain the following result: If \(f\in H^p(\mathbb R^n)\) then there exists linear operators \(S_f: \mathfrak C_\alpha(\mathbb R^n)\to L^1(\mathbb R^n)\) and \(T_f:\mathfrak C_\alpha(\mathbb R^n)\to H^p_{\omega_p}(\mathbb R^n)\) such that \(fg= S_f(g)+T_f(g)\) for any \(g\in \mathfrak C_\alpha(\mathbb R^n)\), where \(\omega_p(x)=1/ (1+|x|)^{n(1-p)}\) when \(n(1/p-1)\notin \mathbb N\) and \(\omega_p(x)=1/ (1+|x|)^{n(1-p)}[\log(e+|x|)]^p\) when \(n(1/p-1)\in \mathbb N\). They also found the corresponding Musielak-Orlicz type space \(\mathcal Y= H^{\Phi_p}(\mathbb R^n)\) in such a way that the original problem was solved. They use a technique of normalization of products of functions via wavelets introduced by \textit{R. R. Coifman} et al. [in: Essays on Fourier analysis in honor of Elias M. Stein. Proceedings of the Princeton conference in harmonic analysis held at Princeton University, Princeton, NJ, May 13-17, 1991 in honor of Elias M. Stein's sixtieth birthday. Princeton, NJ: Princeton Univ. Press. 146--161 (1995; Zbl 0842.42011)].