Quantitative Stability in Fractional Hardy-Sobolev Inequalities: The Role of Euler-Lagrange Equations

Souptik Chakraborty 1 ​ {}^{1\lx@orcidlink{0009-0004-1867-0560}{\lx@orcidlogo}} and Utsab Sarkar 2 ​ {}^{2\lx@orcidlink{0000-0003-4343-0724}{\lx@orcidlogo}} · 2025

Read Quantitative Stability in Fractional Hardy-Sobolev Inequalities: The Role of Euler-Lagrange Equations by Souptik Chakraborty 1 ​ {}^{1\lx@orcidlink{0009-0004-1867-0560}{\lx@orcidlogo}} and Utsab Sarkar 2 ​ {}^{2\lx@orcidlink{0000-0003-4343-0724}{\lx@orcidlogo}} (2025) 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-2503-06716.txt.

Introduction

Souptik Chakraborty and Utsab Sarkar study how stable a fractional Hardy-Sobolev inequality is. Such inequalities bound a weighted size of a function by its fractional energy. The best constant in the bound is known, but how near a function must sit to the best case was open. The work centers on the Euler-Lagrange equation tied to the bound, near a positive ground state.

They prove sharp counts of how far near-best functions can stray. A softer result sorts the near-miss sequences of the linked energy in full. The findings extend classic stability results for the Sobolev bound to fractional calculus. Past work on the matching exponent case is also carried further. The proofs lean on the structure of the equation itself.

Catalog record

CREATORSouptik Chakraborty 1 ​ {}^{1\lx@orcidlink{0009-0004-1867-0560}{\lx@orcidlogo}} and Utsab Sarkar 2 ​ {}^{2\lx@orcidlink{0000-0003-4343-0724}{\lx@orcidlogo}}
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE107 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2503-06716

Browse the full library · About the Tower of Babel

Full text

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

Quantitative Stability in Fractional Hardy-Sobolev Inequalities: The Role of Euler-Lagrange Equations Souptik Chakraborty 1 ​ {}^{1\lx@orcidlink{0009-0004-1867-0560}{\lx@orcidlogo}} and Utsab Sarkar 2 ​ {}^{2\lx@orcidlink{0000-0003-4343-0724}{\lx@orcidlogo}} Abstract . This paper investigates sharp stability estimates for the fractional Hardy-Sobolev inequality: μ s , t ​ ( ℝ N ) ​ ( ∫ ℝ N | u | 2 s ∗ ​ ( t ) | x | t ​ 𝑑 x ) 2 2 s ∗ ​ ( t ) ≤ ∫ ℝ N | ( − Δ ) s 2 ​ u | 2 ​ 𝑑 x , for all ​ u ∈ H ˙ s ​ ( ℝ N ) \mu_{s,t}\left(\mathbb{R}^{N}\right)\left(\int_{\mathbb{R}^{N}}\frac{|u|^{2^{*}_{s}(t)}}{|x|^{t}}\,{\rm d}x\right)^{\frac{2}{2^{*}_{s}(t)}}\leq\int_{\mathbb{R}^{N}}\left|(-\Delta)^{\frac{s}{2}}u\right|^{2}\,{\rm d}x,\quad\text{for all }u\in\dot{H}^{s}\left(\mathbb{R}^{N}\right) where s ∈ ( 0 , 1 ) s\in(0,1) , 0 < t < 2 ​ s 0<t<2s , N > 2 ​ s N>2s is an integer, and 2 s ∗ ​ ( t ) = 2 ​ ( N − t ) N − 2 ​ s 2^{*}_{s}(t)=\frac{2(N-t)}{N-2s} . Here, μ s , t ​ ( ℝ N ) \mu_{s,t}\left(\mathbb{R}^{N}\right) represents the best constant in the inequality. The primary focus is on the quantitative stability results of the above inequality and the corresponding Euler-Lagrange equation near a positive ground-state solution. Additionally, a qualitative stability result is established for the Euler-Lagrange equation, offering a thorough characterization of the Palais-Smale sequences for the associated energy functional. These results generalize the sharp quantitative stability results for the classical Sobolev inequality in ℝ N \mathbb{R}^{N} , originally obtained by Bianchi and Egnell [J. Funct. Anal. 1991] as well as the corresponding critical exponent problem in ℝ N \mathbb{R}^{N} , explored by Ciraolo, Figalli, and Maggi [Int. Math. Res. Not. 2017] in the framework of fractional calculus. 2020 Mathematics Subject Classification 35A23, 35R11, 35B33, 35B35 1. Introduction In recent years, there has been a growing interest in the sharp quantitative stability of various functional and geometric inequalities and their applications in fields such as the calculus of variations, geometric analysis and diffusion flows, among others. Notable examples include the Euclidean isoperimetric inequality, Sobolev inequality and its nonlocal variant, Gagliardo-Nirenberg-Sobolev inequality, Caffarelli-Kohn-Nirenberg inequality and the Poincaré-Sobolev inequality in hyperbolic space, along with many others from a wide range of studies [ 6 , 11 , 9 , 12 , 5 , 7 , 19 , 30 ] . This paper examines the sharp quantitative stability of the fractional Hardy-Sobolev inequality on \mathbb{R}^{N} , beginning with an approach inspired by the work of Bianchi and Egnell [ 6 ] and subsequently extending the analysis to quantify also the sharp stability for the associated Euler-Lagrange equation by following the footsteps of Ciraolo, Figalli, and Maggi [ 13 ] . Our starting point is the following fractional Hardy-Sobolev inequality: (1.1) \mu_{s,t}(\mathbb{R}^{N})\left(\int_{\mathbb{R}^{N}}\frac{|u|^{2^{*}_{s}(t)}}{|x|^{t}}\,\mathrm{d}x\right)^{\frac{2}{2^{*}_{s}(t)}}\leq\int_{\mathbb{R}^{N}}|(-\Delta)^{\frac{s}{2}}u|^{2}\,\mathrm{d}x\hskip 8.53581pt;\text{ for all }u\in\dot{H}^{s}\left(\mathbb{R}^{N}\right) where s\in(0,1) , 0<t<2s<N and 2^{*}_{s}(t)\coloneqq\frac{2(N-t)}{N-2s} . Clearly, 2<2^{*}_{s}(t)<2^{*}_{s}\coloneqq\frac{2N}{N-2s} . The best constant \mu_{s,t}(\mathbb{R}^{N})>0 in ( 1.1 ) is defined by (1.2) \mu_{s,t}(\mathbb{R}^{N})\coloneqq\mu_{s,t}=\inf_{u\in\dot{H}^{s}\left(\mathbb{R}^{N}\right)\setminus\{0\}}\frac{\|u\|_{\dot{H}^{s}\left(\mathbb{R}^{N}\right)}^{2}}{\left\||x|^{-\frac{t}{2_{s}^{*}(t)}}u\right\|_{2_{s}^{*}(t)}^{2}} and is attained by a non-negative, radially symmetric, strictly decreasing function U_{s,t} (see e.g., [ 31 , 22 ] ) such that \||x|^{\frac{-t}{2^{*}_{s}(t)}}U_{s,t}\|_{2^{*}_{s}(t)}=1 . Clearly then U_{s,t} is a weak solution to the Euler-Lagrange (fractional Hardy-Sobolev) equation (1.3) \begin{cases}(-\Delta)^{s}u=\mu_{s,t}\frac{|u|^{2_{s}^{*}(t)-2}u}{|x|^{t}}\quad;u\in\dot{H}^{s}\left(\mathbb{R}^{N}\right)\vskip 2.84526pt\\ u>0\text{ in }\mathbb{R}^{N}\end{cases} associated with ( 1.1 ). Let us fix this minimizer (bubble) U_{s,t} of ( 1.2 ) from now on. By a direct computation, it’s easy to see that for any \lambda>0 , the radial function U^{\lambda}_{s,t}(x)\coloneqq\lambda^{\frac{N-2s}{2}}U_{s,t}\left(\lambda x\right) and its scalar multiples achieve the best constant \mu_{s,t} . Furthermore, for any \lambda>0 , according to our normalization, it holds that \|U^{\lambda}_{s,t}\|_{\dot{H}^{s}}=\|U_{s,t}\|_{\dot{H}^{s}}=\sqrt{\mu_{s,t}}\quad\mbox{ and }\quad\||x|^{\frac{-t}{2^{*}_{s}(t)}}U_{s,t}^{\lambda}\|_{2^{*}_{s}(t)}=\||x|^{\frac{-t}{2^{*}_{s}(t)}}U_{s,t}\|_{2^{*}_{s}(t)}=1. In [ 23 ] , Musina and Nazarov provided a complete classification of the first two eigenspaces corresponding to the eigenvalue problem for the linearized operator \mathcal{L}_{s,t}^{\lambda} (see Appendix A for details) around a positive weak solution U^{\lambda}_{s,t} of the equation ( 1.3 ). Additionally, they established various regularity, uniqueness, and decay properties of these weak solutions. Since these results will be frequently used in this paper, for the reader’s convenience, we paraphrase them here in the following two theorems. Theorem 1.1 (Non-degeneracy [ 23 , Theorem 1.2 and Lemma 4.2] ) . If v\in\dot{H}^{s}\left(\mathbb{R}^{N}\right) is a weak solution of (-\Delta)^{s}v=\mu_{s,t}(2_{s}^{*}(t)-1)\frac{\left(U_{s,t}^{\lambda_{0}}\right)^{2^{*}_{s}(t)-2}}{|x|^{t}}v\hskip 5.69054pt\mbox{ in }\mathbb{R}^{N}, then v=c\partial_{\lambda}\Big|_{\lambda=\lambda_{0}}U_{s,t}^{\lambda} for some c\in{\mathbb{R}}\setminus\{0\} . Moreover, the first two eigenvalues of the linearized operator \mathcal{L}_{s,t}^{\lambda_{0}} ( see ( A.1 ) for definition ) are \eta_{1}=1,\,\eta_{2}=2^{*}_{s}(t)-1 respectively. Theorem 1.2 (Regularity, decay estimates and uniqueness [ 23 , Theorem 1.1] ) . If u\in\dot{H}^{s}\left(\mathbb{R}^{N}\right) is a solution to ( 4.2 ) then u=V^{\lambda}_{s,t}=\mu_{s,t}^{\frac{1}{2^{*}_{s}(t)-2}}U^{\lambda}_{s,t} for some \lambda\in{\mathbb{R}}^{+} . Furthermore, V_{s,t}\in\mathcal{C}^{\infty}\left(\mathbb{R}^{N}\setminus\{0\}\right)\cap L^{\infty}\left(\mathbb{R}^{N}\right)\cap\mathcal{C}^{\alpha}\left(\mathbb{R}^{N}\right) for any \alpha\in[0,2s-t) ; also there exist c,\,C>0 ( depending on V_{s,t}) such that \frac{c}{1+|x|^{N-2s}}\leq V_{s,t}(x)\leq\frac{C}{1+|x|^{N-2s}} for any x\in\mathbb{R}^{N} . For u\in\dot{H}^{s}\left(\mathbb{R}^{N}\right) , we consider the deficit functional corresponding to the inequality ( 1.1 ) (1.4) \delta(u)\coloneqq\|u\|_{\dot{H}^{s}\left(\mathbb{R}^{N}\right)}^{2}-\mu_{s,t}\left\||x|^{-\frac{t}{2_{s}^{*}(t)}}u\right\|_{2_{s}^{*}(t)}^{2}. Now from the above discussion, it is immediate that the set where the equality holds for ( 1.1 ), i.e., when \delta(u)=0 , is precisely the two-dimensional smooth manifold \mathcal{M}\subset\dot{H}^{s}\left(\mathbb{R}^{N}\right) defined as (1.5) \mathcal{M}\coloneqq\left\{cU_{s,t}^{\lambda}\,\colon\,(c,\lambda)\in{\mathbb{R}}\times{\mathbb{R}}^{+},\,U_{s,t}^{\lambda}(x)=\lambda^{\frac{N-2s}{2}}U_{s,t}(\lambda x)\right\}. The question of stability asks whether, for a given u\in\dot{H}^{s}\left(\mathbb{R}^{N}\right) , a sufficiently small \delta(u) implies that u is close to \mathcal{M} in the \dot{H}^{s}\left(\mathbb{R}^{N}\right) topology. Moreover, can one determine how \delta(u) controls the \dot{H}^{s}\left(\mathbb{R}^{N}\right) distance between u and \mathcal{M} in a quantitative manner? Brézis and Lieb [ 8 ] originally posed this question in the context of the classical Sobolev inequality, where the extremal functions were characterized by those so-called Aubin-Talenti bubbles [ 2 , 28 ] . Now for the fractional Sobolev inequality (i.e., when 0<2s<N ), the minimizers were classified by Cotsiolis et al. [ 14 ] . The first in-depth study on the sharp stability on the classical Sobolev inequality was conducted in the seminal work of Bianchi and Egnell [ 6 ] , as well as by Rey [ 25 ] . Later, the stability result established by Bianchi and Egnell was extended to biharmonic operators in [ 21 ] and further generalized to poly-harmonic operators by Bartsch et al. [ 3 ] . The sharp stability for the fractional Sobolev inequality, in the case where 0<2s<N , was subsequently investigated by Chen et al. [ 11 ] . Turning our attention from inequality to the derivation of the corresponding Euler-Lagrange equation, we examine the first variation of the deficit functional \delta(\cdot) in the vicinity of a minimizer u of \mu_{s,t} . i.e., \frac{{\rm d}}{{\rm d}t}\Big|_{t=0}\delta(u+t\phi)=0 for all \phi\in\dot{H}^{s}\left(\mathbb{R}^{N}\right) , under the constraint \left\||x|^{-\frac{t}{2_{s}^{*}(t)}}u\right\|_{2_{s}^{*}(t)}=1 . As a result, we obtain ( 1.3 ) without imposing any sign restriction on u . In other words, the set of all critical points of the energy functional in ( 3.7 ) includes all minimizers of \mu_{s,t} . However, Theorem 1.2 , which characterizes all solutions of ( 1.3 ) that do not change sign, guarantees that the sub-manifold \mathcal{M}_{1}\coloneqq\{cU_{s,t}^{\lambda}\mid(c,\lambda)\in\{\pm 1\}\times\mathbb{R}^{+}\} precisely corresponds to these non-sign-changing critical points. For u\in\dot{H}^{s}\left(\mathbb{R}^{N}\right) , consider the deficit functional corresponding to ( 4.2 ) as (1.6) \Gamma(u)\coloneqq\left\|(-\Delta)^{s}u-\frac{|u|^{2_{s}^{*}(t)-2}u}{|x|^{t}}\right\|_{\left(\dot{H}^{s}\left(\mathbb{R}^{N}\right)\right)^{\prime}}=\left\|I_{s,t}^{\prime}(u)\right\|_{\left(\dot{H}^{s}\left(\mathbb{R}^{N}\right)\right)^{\prime}}, where I_{s,t} is defined in ( 4.1 ). We begin by exploring the qualitative stability question in this context, specifically, whether a sufficiently small value of \Gamma(u) for some non-sign-changing u\in\dot{H}^{s}\left(\mathbb{R}^{N}\right) implies that u is close to the sub-manifold \mathcal{M}_{1} of non-sign-changing critical points of I_{s,t} . The answer to this question is

Read the complete text