{"title":"Eigenvalue problems in 𝐿^{∞}: optimality conditions, duality, and relations with optimal transport","authors":"Leon Bungert, Yury Korolev","doi":"10.1090/cams/11","DOIUrl":null,"url":null,"abstract":"<p>In this article we characterize the <inline-formula content-type=\"math/mathml\">\n<mml:math xmlns:mml=\"http://www.w3.org/1998/Math/MathML\" alttext=\"normal upper L Superscript normal infinity\">\n <mml:semantics>\n <mml:msup>\n <mml:mrow class=\"MJX-TeXAtom-ORD\">\n <mml:mi mathvariant=\"normal\">L</mml:mi>\n </mml:mrow>\n <mml:mi mathvariant=\"normal\">∞<!-- ∞ --></mml:mi>\n </mml:msup>\n <mml:annotation encoding=\"application/x-tex\">\\mathrm {L}^\\infty</mml:annotation>\n </mml:semantics>\n</mml:math>\n</inline-formula> eigenvalue problem associated to the Rayleigh quotient <inline-formula content-type=\"math/mathml\">\n<mml:math xmlns:mml=\"http://www.w3.org/1998/Math/MathML\" alttext=\"double-vertical-bar nabla u double-vertical-bar Subscript normal upper L Sub Superscript normal infinity Baseline slash double-vertical-bar u double-vertical-bar Subscript normal infinity\">\n <mml:semantics>\n <mml:mrow>\n <mml:mo fence=\"true\" stretchy=\"true\" symmetric=\"true\" />\n <mml:mrow class=\"MJX-TeXAtom-ORD\">\n <mml:mo fence=\"false\" stretchy=\"false\">‖<!-- ‖ --></mml:mo>\n <mml:mi mathvariant=\"normal\">∇<!-- ∇ --></mml:mi>\n <mml:mi>u</mml:mi>\n <mml:msub>\n <mml:mo fence=\"false\" stretchy=\"false\">‖<!-- ‖ --></mml:mo>\n <mml:mrow class=\"MJX-TeXAtom-ORD\">\n <mml:msup>\n <mml:mrow class=\"MJX-TeXAtom-ORD\">\n <mml:mi mathvariant=\"normal\">L</mml:mi>\n </mml:mrow>\n <mml:mi mathvariant=\"normal\">∞<!-- ∞ --></mml:mi>\n </mml:msup>\n </mml:mrow>\n </mml:msub>\n </mml:mrow>\n <mml:mrow class=\"MJX-TeXAtom-CLOSE\">\n\n </mml:mrow>\n <mml:mo stretchy=\"true\">/</mml:mo>\n <mml:mrow class=\"MJX-TeXAtom-OPEN\">\n\n </mml:mrow>\n <mml:mrow class=\"MJX-TeXAtom-ORD\">\n <mml:mo fence=\"false\" stretchy=\"false\">‖<!-- ‖ --></mml:mo>\n <mml:mi>u</mml:mi>\n <mml:msub>\n <mml:mo fence=\"false\" stretchy=\"false\">‖<!-- ‖ --></mml:mo>\n <mml:mi mathvariant=\"normal\">∞<!-- ∞ --></mml:mi>\n </mml:msub>\n </mml:mrow>\n <mml:mo fence=\"true\" stretchy=\"true\" symmetric=\"true\" />\n </mml:mrow>\n <mml:annotation encoding=\"application/x-tex\">\\left .{\\|\\nabla u\\|_{\\mathrm {L}^\\infty }}\\middle /{\\|u\\|_\\infty }\\right .</mml:annotation>\n </mml:semantics>\n</mml:math>\n</inline-formula> and relate it to a divergence-form PDE, similarly to what is known for <inline-formula content-type=\"math/mathml\">\n<mml:math xmlns:mml=\"http://www.w3.org/1998/Math/MathML\" alttext=\"normal upper L Superscript p\">\n <mml:semantics>\n <mml:msup>\n <mml:mrow class=\"MJX-TeXAtom-ORD\">\n <mml:mi mathvariant=\"normal\">L</mml:mi>\n </mml:mrow>\n <mml:mi>p</mml:mi>\n </mml:msup>\n <mml:annotation encoding=\"application/x-tex\">\\mathrm {L}^p</mml:annotation>\n </mml:semantics>\n</mml:math>\n</inline-formula> eigenvalue problems and the <inline-formula content-type=\"math/mathml\">\n<mml:math xmlns:mml=\"http://www.w3.org/1998/Math/MathML\" alttext=\"p\">\n <mml:semantics>\n <mml:mi>p</mml:mi>\n <mml:annotation encoding=\"application/x-tex\">p</mml:annotation>\n </mml:semantics>\n</mml:math>\n</inline-formula>-Laplacian for <inline-formula content-type=\"math/mathml\">\n<mml:math xmlns:mml=\"http://www.w3.org/1998/Math/MathML\" alttext=\"p greater-than normal infinity\">\n <mml:semantics>\n <mml:mrow>\n <mml:mi>p</mml:mi>\n <mml:mo>></mml:mo>\n <mml:mi mathvariant=\"normal\">∞<!-- ∞ --></mml:mi>\n </mml:mrow>\n <mml:annotation encoding=\"application/x-tex\">p>\\infty</mml:annotation>\n </mml:semantics>\n</mml:math>\n</inline-formula>. Contrary to existing methods, which study <inline-formula content-type=\"math/mathml\">\n<mml:math xmlns:mml=\"http://www.w3.org/1998/Math/MathML\" alttext=\"normal upper L Superscript normal infinity\">\n <mml:semantics>\n <mml:msup>\n <mml:mrow class=\"MJX-TeXAtom-ORD\">\n <mml:mi mathvariant=\"normal\">L</mml:mi>\n </mml:mrow>\n <mml:mi mathvariant=\"normal\">∞<!-- ∞ --></mml:mi>\n </mml:msup>\n <mml:annotation encoding=\"application/x-tex\">\\mathrm {L}^\\infty</mml:annotation>\n </mml:semantics>\n</mml:math>\n</inline-formula>-problems as limits of <inline-formula content-type=\"math/mathml\">\n<mml:math xmlns:mml=\"http://www.w3.org/1998/Math/MathML\" alttext=\"normal upper L Superscript p\">\n <mml:semantics>\n <mml:msup>\n <mml:mrow class=\"MJX-TeXAtom-ORD\">\n <mml:mi mathvariant=\"normal\">L</mml:mi>\n </mml:mrow>\n <mml:mi>p</mml:mi>\n </mml:msup>\n <mml:annotation encoding=\"application/x-tex\">\\mathrm {L}^p</mml:annotation>\n </mml:semantics>\n</mml:math>\n</inline-formula>-problems for <inline-formula content-type=\"math/mathml\">\n<mml:math xmlns:mml=\"http://www.w3.org/1998/Math/MathML\" alttext=\"p right-arrow normal infinity\">\n <mml:semantics>\n <mml:mrow>\n <mml:mi>p</mml:mi>\n <mml:mo stretchy=\"false\">→<!-- → --></mml:mo>\n <mml:mi mathvariant=\"normal\">∞<!-- ∞ --></mml:mi>\n </mml:mrow>\n <mml:annotation encoding=\"application/x-tex\">p\\to \\infty</mml:annotation>\n </mml:semantics>\n</mml:math>\n</inline-formula>, we develop a novel framework for analyzing the limiting problem directly using convex analysis and geometric measure theory. For this, we derive a novel fine characterization of the subdifferential of the Lipschitz-constant-functional <inline-formula content-type=\"math/mathml\">\n<mml:math xmlns:mml=\"http://www.w3.org/1998/Math/MathML\" alttext=\"u right-arrow from bar double-vertical-bar nabla u double-vertical-bar Subscript normal upper L Sub Superscript normal infinity\">\n <mml:semantics>\n <mml:mrow>\n <mml:mi>u</mml:mi>\n <mml:mo stretchy=\"false\">↦<!-- ↦ --></mml:mo>\n <mml:mo fence=\"false\" stretchy=\"false\">‖<!-- ‖ --></mml:mo>\n <mml:mi mathvariant=\"normal\">∇<!-- ∇ --></mml:mi>\n <mml:mi>u</mml:mi>\n <mml:msub>\n <mml:mo fence=\"false\" stretchy=\"false\">‖<!-- ‖ --></mml:mo>\n <mml:mrow class=\"MJX-TeXAtom-ORD\">\n <mml:msup>\n <mml:mrow class=\"MJX-TeXAtom-ORD\">\n <mml:mi mathvariant=\"normal\">L</mml:mi>\n </mml:mrow>\n <mml:mi mathvariant=\"normal\">∞<!-- ∞ --></mml:mi>\n </mml:msup>\n </mml:mrow>\n </mml:msub>\n </mml:mrow>\n <mml:annotation encoding=\"application/x-tex\">u\\mapsto \\|\\nabla u\\|_{\\mathrm {L}^\\infty }</mml:annotation>\n </mml:semantics>\n</mml:math>\n</inline-formula>. We show that the eigenvalue problem takes the form <inline-formula content-type=\"math/mathml\">\n<mml:math xmlns:mml=\"http://www.w3.org/1998/Math/MathML\" alttext=\"lamda nu u equals minus d i v left-parenthesis tau nabla Subscript tau Baseline u right-parenthesis\">\n <mml:semantics>\n <mml:mrow>\n <mml:mi>λ<!-- λ --></mml:mi>\n <mml:mi>ν<!-- ν --></mml:mi>\n <mml:mi>u</mml:mi>\n <mml:mo>=</mml:mo>\n <mml:mo>−<!-- − --></mml:mo>\n <mml:mi>div</mml:mi>\n <mml:mo><!-- --></mml:mo>\n <mml:mo stretchy=\"false\">(</mml:mo>\n <mml:mi>τ<!-- τ --></mml:mi>\n <mml:msub>\n <mml:mi mathvariant=\"normal\">∇<!-- ∇ --></mml:mi>\n <mml:mi>τ<!-- τ --></mml:mi>\n </mml:msub>\n <mml:mi>u</mml:mi>\n <mml:mo stretchy=\"false\">)</mml:mo>\n </mml:mrow>\n <mml:annotation encoding=\"application/x-tex\">\\lambda \\nu u =-\\operatorname {div}(\\tau \\nabla _\\tau u)</mml:annotation>\n </mml:semantics>\n</mml:math>\n</inline-formula>, where <inline-formula content-type=\"math/mathml\">\n<mml:math xmlns:mml=\"http://www.w3.org/1998/Math/MathML\" alttext=\"nu\">\n <mml:semantics>\n <mml:mi>ν<!-- ν --></mml:mi>\n <mml:annotation encoding=\"application/x-tex\">\\nu</mml:annotation>\n </mml:semantics>\n</mml:math>\n</inline-formula> and <inline-formula content-type=\"math/mathml\">\n<mml:math xmlns:mml=\"http://www.w3.org/1998/Math/MathML\" alttext=\"tau\">\n <mml:semantics>\n <mml:mi>τ<!-- τ --></mml:mi>\n <mml:annotation encoding=\"application/x-tex\">\\tau</mml:annotation>\n </mml:semantics>\n</mml:math>\n</inline-formula> are non-negative measures concentrated where <inline-formula content-type=\"math/mathml\">\n<mml:math xmlns:mml=\"http://www.w3.org/1998/Math/MathML\" alttext=\"StartAbsoluteValue u EndAbsoluteValue\">\n <mml:semantics>\n <mml:mrow>\n <mml:mrow class=\"MJX-TeXAtom-ORD\">\n <mml:mo stretchy=\"false\">|</mml:mo>\n </mml:mrow>\n <mml:mi>u</mml:mi>\n <mml:mrow class=\"MJX-TeXAtom-ORD\">\n <mml:mo stretchy=\"false\">|</mml:mo>\n </mml:mrow>\n </mml:mrow>\n <mml:annotation encoding=\"application/x-tex\">|u|</mml:annotation>\n </mml:semantics>\n</mml:math>\n</inline-formula> respectively <inline-formula content-type=\"math/mathml\">\n<mml:math xmlns:mml=\"http://www.w3.org/1998/Math/MathML\" alttext=\"StartAbsoluteValue nabla u EndAbsoluteValue\">\n <mml:semantics>\n <mml:mrow>\n <mml:mrow class=\"MJX-TeXAtom-ORD\">\n <mml:mo stretchy=\"false\">|</mml:mo>\n </mml:mrow>\n <mml:mi mathvariant=\"normal\">∇<!-- ∇ --></mml:mi>\n <mml:mi>u</mml:mi>\n <mml:mrow class=\"MJX-TeXAtom-ORD\">\n <mml:mo stretchy=\"false\">|</mml:mo>\n </mml:mrow>\n </mml:mrow>\n <mml:annotation encoding=\"application/x-tex\">|\\nabla u|</mml:annotation>\n </mml:semantics>\n</mml:math>\n</inline-formula> are maximal, and <inline-formula content-type=\"math/mathml\">\n<mml:math xmlns:mml=\"http://www.w3.org/1998/Math/MathML\" alttext=\"nabla Subscript tau Baseline u\">\n <mml:semantics>\n <mml:mrow>\n <mml:msub>\n <mml:mi mathvariant=\"normal\">∇<!-- ∇ --></mml:mi>\n <mml:mi>τ<!-- τ --></mml:mi>\n </mml:msub>\n <mml:mi>u</mml:mi>\n </mml:mrow>\n <mml:annotation encoding=\"application/x-tex\">\\nabla _\\tau u</mml:annotation>\n </mml:semantics>\n</mml:math>\n</inline-formula> is the tangential gradient of <inline-formula content-type=\"math/mathml\">\n<mml:math xmlns:mml=\"http://www.w3.org/1998/Math/MathML\" alttext=\"u\">\n <mml:semantics>\n <mml:mi>u</mml:mi>\n <mml:annotation encoding=\"application/x-tex\">u</mml:annotation>\n </mml:semantics>\n</mml:math>\n</inline-formula>","PeriodicalId":285678,"journal":{"name":"Communications of the American Mathematical Society","volume":"13 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2021-07-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"6","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Communications of the American Mathematical Society","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1090/cams/11","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 6
Abstract
In this article we characterize the L∞\mathrm {L}^\infty eigenvalue problem associated to the Rayleigh quotient ‖∇u‖L∞/‖u‖∞\left .{\|\nabla u\|_{\mathrm {L}^\infty }}\middle /{\|u\|_\infty }\right . and relate it to a divergence-form PDE, similarly to what is known for Lp\mathrm {L}^p eigenvalue problems and the pp-Laplacian for p>∞p>\infty. Contrary to existing methods, which study L∞\mathrm {L}^\infty-problems as limits of Lp\mathrm {L}^p-problems for p→∞p\to \infty, we develop a novel framework for analyzing the limiting problem directly using convex analysis and geometric measure theory. For this, we derive a novel fine characterization of the subdifferential of the Lipschitz-constant-functional u↦‖∇u‖L∞u\mapsto \|\nabla u\|_{\mathrm {L}^\infty }. We show that the eigenvalue problem takes the form λνu=−div(τ∇τu)\lambda \nu u =-\operatorname {div}(\tau \nabla _\tau u), where ν\nu and τ\tau are non-negative measures concentrated where |u||u| respectively |∇u||\nabla u| are maximal, and ∇τu\nabla _\tau u is the tangential gradient of uu