Best Constant and extremal functions for a class HARDY-SOBOLEV-MAZ’YA INEQUALITIES

Daowen Lin, Xi-Nan Ma · 2024

Read Best Constant and extremal functions for a class HARDY-SOBOLEV-MAZ’YA INEQUALITIES by Daowen Lin, Xi-Nan Ma (2024) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2412-09033.txt.

Introduction

Daowen Lin and Xi-Nan Ma pin down the best constant in a family of Hardy-Sobolev-Maz'ya inequalities, which blend different ways of measuring a function's size. Their route runs through a p-Laplace equation, a nonlinear cousin of the classic Laplace equation. They first derive an integral identity for that equation. With careful prior estimates, they then classify every positive finite-energy solution that is symmetric around a cylinder, across a middle range of dimensions.

An earlier result by Secchi, Smets, and Willem completes the chain. Together, the pieces yield both the sharp constant and the extremal functions that attain it.

Catalog record

CREATORDaowen Lin, Xi-Nan Ma
YEAR2024
COLLECTIONPAPERS
FORMATTXT
FILE SIZE45 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2412-09033

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Best Constant and extremal functions for a class HARDY-SOBOLEV-MAZ’YA INEQUALITIES Daowen Lin, Xi-Nan Ma Abstract . We derive an integral identity for a class p p -Laplace equation, and then classify all positive finite energy cylindrically symmetric solutions of the equation ( 1.2 ) for 3 ≤ k ≤ n − 1 , 3\leq k\leq n-1, with the help of some a prior estimates. Combining this with the result of Secchi-Smets-Willem [ 22 ] , as a consequence, we obtain the best constant and extremal functions for the related Hardy-Sobolev-Maz’ya inequalities. 1. Introduction In this paper, we study the positive extremals for the following Hardy-Sobolev-Maz’ya inequalities when s=1,k\geq 3 . Theorem 1.1 . [ 4 ] Let n\geq 3 , 2\leq k\leq n , and p,s be real numbers satisfying 1<p<n , 0\leq s\leq p and s<k , There exists a positive constant S_{p,s}=S(p,s,n,k) such that for all u\in D^{1,p}(\mathbb{R}^{n}) , we have (1.1) \displaystyle\left(\int_{\mathbb{R}^{n}}\frac{|u|^{\frac{p(n-s)}{n-p}}}{|y|^{s}}dx\right)^{\frac{n-p}{p(n-s)}}\leq S_{p,s}\left(\int_{\mathbb{R}^{n}}|\nabla u|^{p}dx\right)^{\frac{1}{p}}, where for p>1 , we define the space D^{1,p}(\mathbb{R}^{n}) as the closure of C_{c}^{\infty}(\mathbb{R}^{n}) with respect to the norm \displaystyle||u||_{D^{1,p}(\mathbb{R}^{n})}=\left(\int_{\mathbb{R}^{n}}|\nabla u|^{p}dx\right)^{\frac{1}{p}}. For s=1 , the Euler-Lagrange equation associated to ( 1.1 ) is (1.2) \displaystyle-\Delta_{p}u \displaystyle=\frac{u^{p^{*}(1)-1}}{|y|}\,\,\textrm{in}\,\,\mathbb{R}^{n}, \displaystyle u \displaystyle>0, \displaystyle u \displaystyle\in D^{1,p}(\mathbb{R}^{n}), where p^{*}(1)=\frac{p(n-1)}{n-p} , x=(y,z) , \mathbb{R}^{n}=\mathbb{R}^{k}\times\mathbb{R}^{n-k} . If s=p , ( 1.1 ) is an extension of the classical Hardy inequalities. The case s=0 corresponds to the classical Sobolev inequalities, whose finite energy extremal functions have been classified by [ 21 ] and [ 26 ] , or see [ 12 ] for the convex cone case via integral by part method and a priori estimates. In more general case 0\leq s<p and k=n , the best constant have been computed in [ 17 ] , their extremal functions are, up to dilations and translations, \displaystyle U(x)=(1+|x|^{\frac{p-s}{p-1}})^{-\frac{n-p}{p-s}} More general inequalities have been considered by Caffarelli-Kohn-Nirenberg [ 6 ] : \displaystyle\left(\int_{\mathbb{R}^{d}}|x|^{-bq}u^{q}dx\right)^{\frac{1}{q}}\leq C_{a,b}\left(\int_{\mathbb{R}^{d}}|x|^{-ap}|\nabla u|^{p}dx\right)^{\frac{1}{p}} where p>1,a<b<a+1,q=\frac{dp}{d-p(1+a-b)},0\leq a<\frac{d}{p}-1,C_{a,b}=C_{a,b}(d,p,a,b,q). They turn out to be of the same form as in the case a=0 , for general, we refer to [ 14 ] and [ 11 ] for the recent progress. When k<n , extremal functions cannot be anymore radially symmetric. Cylindrical symmetry of some extremal functions has been proved Secchi-Smets-Willem [ 22 ] , by symmetrizations arguments. For p=2 , s=1 , Mancini-Fabbri-Sandeep [ 19 ] classified all positive finite energy solutions using moving plane method and a differential identity from [ 16 ] with some a prior pointwise estimates. Independently, Alvino-Ferone-Trombetti also investigated when s=1 in [ 1 ] , they have a conjecture as follow in their article, but they could only address when n=3,k=2,p=2 . Conjecture 1 . [ 1 ] Let u_{0} be the function given by (1.3) \displaystyle u_{0}(x)=u_{0}(y,z)=C_{p,n,k}[(1+|y|)^{2}+|z|^{2}]^{-\frac{n-p}{2(p-1)}}, where C_{p,n,k}=(k-1)^{\frac{n-p}{p(p-1)}}(n-p)^{\frac{n-p}{p}}(p-1)^{-\frac{n-p}{p}} . Then u is an extremal function for ( 1.1 ) if u(y,z)=\lambda^{\frac{n-p}{p}}u_{0}(\lambda y,\lambda z+z_{0}) for some \lambda>0 and z_{0}\in\mathbb{R}^{n-k} . Invoking the cylindrical symmetry results in [ 22 ] , we generalize the differential identity in [ 16 ] , which is motivated from the idea by Jerison-Lee [ 18 ] . But unfortunately, we can’t get the similar a prior point estimates for p -Laplace equations. Since what we need to characterize the cylindrically symmetric solutions of ( 1.2 ) are almost integral estimates, with the help of sharp C^{0} estimate in [ 15 ] , the gradient estimates using the ideas from Xiang [ 27 ] and Antonini-Ciraolo-Farina [ 2 ] . We can tackle some cases: Theorem 1.2 . For s=1,n\geq 4,3\leq k\leq n-1 , u is a cylindrically symmetric solution of ( 1.2 ) then u(y,z)=\lambda^{\frac{n-p}{p}}u_{0}(\lambda y,\lambda z+z_{0}) for some \lambda>0 and z_{0}\in\mathbb{R}^{n-k} . From the above theorem, we can give a positive answer for the conjecture by Alvino-Ferone-Trombetti [ 1 ] when k\geq 3 . Since we can’t get the a priori estimates, the case k=2 is still open. As we had remarked that the best constant of Hardy-Sobolev-Maz’ya inequalities ( 1.1 ) achieve at some cylindrically symmetric functions [ 22 ] . We give some more reference for the best constant and extremal function on the Hardy-Sobolev-Maz’ya inequalities on p=2 , please see the recent progress in [ [ 5 ] , [ 7 ] , [ 8 ] , [ 9 ] , [ 20 ] , [ 25 ] ] etc., they used the hyperbolic symmetry or the moving plane methods. The paper is organized as follows. In section 2, we first review the cylindrically symmetric for the extremal function by Secchi-Smets-Willem [ 22 ] . Then we give the sharp C^{0} estimate by Dutta [ 15 ] , the gradient estimates using the idea from Xiang [ 27 ] and Antonini-Ciraolo-Farina [ 2 ] . In section 3, we give a generalization of the differential identity for a class p-Laplace operator via the ideas from Serrin-Zou [ 23 ] , [ 16 ] and [ 19 ] , as we noted which is motivated from the idea Jerison-Lee [ 18 ] for the Laplace operator case. Then, using this generalized differential identity and the integral estimates in section 2, we complete the proof of Theorem 1.2 . 2. cylindrically symmetry properties and a priori estimates In this sections, we first show briefly that the best constant of Hardy-Sobolev-Maz’ya inequalities achieve at some cylindrically symmetric functions in Secchi-Smets-Willem [ 22 ] , and we state the C^{0} estimates by Dutta [ 15 ] . Then we mainly obtain the integral gradient estimates in Lemma 2.4 and Lemma 2.5 , where we use the ideas from Xiang [ 27 ] and Antonini-Ciraolo-Farina [ 2 ] . These integral estimates will be used in next section to get the proof of the main theorem. As in [ 22 ] , we consider the minimization problem (2.1) \displaystyle S_{p,s}=S(p,s,n,k)=\inf\left\{\int_{\mathbb{R}^{n}}|\nabla u|^{p}|u\in D^{1,p}(\mathbb{R}^{n}),\int_{\mathbb{R}^{n}}\frac{|u|^{p^{*}(1)}}{|y|}=1\right\}. Let u\in L^{1}(\mathbb{R}^{n}) be a non-negative function. Let us denote by u^{*}(*,z) the Schwarz symmetrization of u(*,z) and u^{**}(y,*) the Schwarz symmetrization of u^{*}(y,*) . It is clear that u^{**} depends only on (|y|,|z|) . Let us define \displaystyle D_{**}^{1,p}(\mathbb{R}^{n})=\{u\in D^{1,p}(\mathbb{R}^{n})|u=u^{**}\}, and (2.2) \displaystyle S^{**}_{p,s}=S^{**}(p,s,n,k)=\inf\left\{\int_{\mathbb{R}^{n}}|\nabla u|^{p}|u\in D_{**}^{1,p}(\mathbb{R}^{n}),\int_{\mathbb{R}^{n}}\frac{|u|^{p^{*}(1)}}{|y|}=1\right\}. From [ 22 ] , first we have the following property. Proposition 2.1 . [ 22 ] S_{p,s}=S^{**}_{p,s} . Then we know the best constant of Hardy-Sobolev-Maz’ya inequalities ( 1.1 ) achieve at some cylindrically symmetric functions. Now we state a priori C^{0} estimates for the Euler-Lagrange equation ( 1.2 ) from Dutta [ 15 ] . Proposition 2.2 . [ 15 ] Let u be a solution of ( 1.2 ), then for any x\in\mathbb{R}^{n} , there are universal constants c,C such that (2.3) \displaystyle\frac{c}{1+|x|^{\frac{n-p}{p-1}}}\leq u(x)\leq\frac{C}{1+|x|^{\frac{n-p}{p-1}}}. We state a a priori gradient estimates for a class p -Laplace equation from Xiang [ 27 ] . Proposition 2.3 . [ 27 ] Let \Omega be a domain in \mathbb{R}^{n} and f\in L^{\infty}_{loc}(\Omega) . Let w\in W^{1,p}_{loc}(\Omega) be a weak solution to equation -\Delta_{p}w=f in \Omega , that is, \int_{\Omega}|\nabla w|^{p-2}\nabla w\cdot\nabla\varphi=\int_{\Omega}f\varphi,\,\,\forall\,\varphi\in C^{\infty}_{0}(\Omega). Then, for any ball B_{2\delta}(x_{0})\subset\Omega , there holds (2.4) \sup_{B_{\frac{\delta}{2}}(x_{0})}|\nabla w|\leq C\left(\delta^{-n}\int_{B_{\delta}(x_{0})}|\nabla w|^{p}\right)^{\frac{1}{p}}+C\delta^{\frac{1}{p-1}}||f||^{\frac{1}{p-1}}_{\infty,B_{\delta}(x_{0})}. We shall use the Proposition 2.2 and Proposition 2.3 to get the following integral estimates ( 2.6 ) when k>2p . Lemma 2.4 . Let u be a solution of ( 1.2 ), for any z\in\mathbb{R}^{n-k} , we have

Read the complete text