On the ( m , 1 ) (m,1) -configuration for the HRT Conjecture
Vignon Oussa · 2025
Read On the ( m , 1 ) (m,1) -configuration for the HRT Conjecture by Vignon Oussa (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-2508-18529.txt.
Introduction
Vignon Oussa attacks a special case of a famous open problem about time-frequency shifts of smooth functions, in a 2025 paper. The HRT conjecture limits how a function and its shifts can be linearly linked. Oussa studies very smooth Schwartz functions obeying one identity: a mix of translates equals a single frequency shift. On the Fourier side, the identity becomes a multiplying rule along a ladder of points.
He rewrites the growth of that product as an average over a compact group, using ergodic theory. If the key average is nonzero, or a recurrence condition holds, the transform must grow exponentially somewhere. Schwartz functions must instead decay fast, so the identity collapses. The result conditionally bars such identities for this configuration.
Catalog record
| CREATOR | Vignon Oussa |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 19 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2508-18529
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
On the ( m , 1 ) (m,1) -configuration for the HRT Conjecture Vignon Oussa Abstract We prove a linear–independence principle for finite families of translates in irreducible, square–integrable modulo the center (SI/Z) representations of step–two nilpotent Lie groups with one–dimensional center. Fix a one–parameter subgroup exp ( ℝ X ) \exp(\mathbb{R}X) and a noncentral element Y Y (not assumed to commute with X X ). For every nonzero smooth vector f f , any finite collection { π ( exp ( x k X ) ) f } k = 1 m \{\pi(\exp(x_{k}X))f\}_{k=1}^{m} (with distinct parameters x k x_{k} ) of translates along exp ( ℝ X ) \exp(\mathbb{R}X) , together with the single translate π ( exp Y ) f \pi(\exp Y)f , is linearly independent. The proof combines Schwartz regularity of matrix coefficients in the noncentral variables with an ergodic log–product estimate on a compact torus generated by commutator phases. As a result, we obtain an HRT–type statement for ( m , 1 ) (m,1) configurations (on–line direction with zero X k 0 X_{k_{0}} –component and rogue exp X k 0 \exp X_{k_{0}} ) for all nonzero Schwartz windows, without spacing restrictions. 1 Motivation and Introduction The Heil–Ramanathan–Topiwala (HRT) conjecture asserts that for any nonzero window f\in L^{2}(\mathbb{R}) and any finite set \Lambda=\{(x_{k},y_{k})\}_{k=1}^{m}\subset\mathbb{R}^{2} , the time–frequency system \mathcal{G}(\Lambda,f):=\big\{\pi(x_{k},y_{k})f:1\leq k\leq m\big\} where \pi(x,y)=M_{y}T_{x} , (M_{y}f)(t)=e^{-2\pi iyt}f(t) , and (T_{x}f)(t)=f(t-x) is linearly independent. The conjecture is corroborated in several important instances. In particular, for collinear configurations (all points lying on a single line in the time–frequency plane), metaplectic reduction carries the problem to pure translations or pure modulations, where linear independence is classical; see, for example, the survey [ 11 ] and related developments in [ 14 , 10 ] . Beyond the collinear case, partial results for two-line patterns are known. The (2,2) configuration is settled in L^{2} by Demeter–Zaharescu [ 5 ] , while Demeter proved that for the (1,3) configuration the conjecture holds for every nonzero Schwartz window [ 4 ] . Moreover, Liu established that HRT holds for almost every (1,3) configuration (with respect to Lebesgue measure on the parameters) for all nonzero L^{2} windows [ 13 ] . Main contribution Motivated by these advances, we study configurations consisting of m points on a line together with one point off that line; an (m,1) pattern in the time–frequency plane. The structural feature making this case tractable is that the on-line points generate a commuting family of time–frequency operators, whereas the single off-line point typically fails to commute with that family. By simultaneously diagonalizing the commuting part, one obtains a functional equation driven by the rogue shift; and the analysis of such an equation leads to new results for this specific configuration. Our approach is mostly representation-theoretic. We work on step-two nilpotent Lie groups with one-dimensional center and employ square-integrable-mod-center (SI/Z) representations. Within this setting, we prove that a finite family consisting of translates along a fixed one-parameter subgroup, together with one rogue translate, is linearly independent for every nonzero smooth vector for the representation at hand. As an immediate corollary, in the Schrödinger realization of the action of the (2d+1) -dimensional Heisenberg group on L^{2}(\mathbb{R}^{d}) , we obtain the HRT-type conclusion for (m,1) configurations without spacing restrictions (as considered in [ 10 ] , for instance) for Schwartz functions. Remark 1.1 (Heisenberg equivalence and acknowledgment) . It is known that any connected and simply connected two-step nilpotent Lie group with a one-dimensional center is Lie isomorphic to a Heisenberg group (Dixmier, Enveloping Algebras , §4.6.1, [ 6 ] ). For generality, we have chosen to formulate our results in general Lie-theoretic terms rather than working directly within a fixed Schrödinger realization. The author thanks Ingrid Beltita for the reference and for pointing out some redundancies (subsequently removed) in earlier drafts. 2 Main theorem Theorem 2.1 . Let N be a connected, simply connected, step–two nilpotent Lie group with one-dimensional center, and let \pi be an irreducible unitary SI/Z (square-integrable modulo the center) representation on a Hilbert space H_{\pi} . Fix a basis \{B_{1}=Z,\ B_{2}=Y_{1},\dots,B_{\ell+1}=Y_{\ell},\ B_{\ell+2}=X\} adapted to the central series of \mathfrak{n}=\mathrm{Lie}(N) , i.e. \mathfrak{n}(k)=\mathrm{span}_{\mathbb{R}}\{B_{1},\dots,B_{k}\} is an ideal in \mathfrak{n} for each k . Let Y\in\mathrm{span}\{Y_{1},\dots,Y_{\ell}\} be nonzero and f\in H_{\pi}^{\infty} be nonzero (a non trival smooth vector.) Then for any distinct real parameters x_{1}<\dots<x_{m}, the finite set \big\{\pi(\exp(x_{k}X))f:1\leq k\leq m\big\}\ \cup\ \big\{\pi(\exp Y)f\big\} is linearly independent in H_{\pi} . 3 Proof of Theorem 2.1 Standing assumptions and notation. Throughout Sections 3.1 – 3.3 we assume: • N is a connected, simply connected, step–two nilpotent Lie group with one–dimensional center Z(N) . • \mathfrak{n}=\mathrm{Lie}(N) with basis \{Z,Y_{1},\dots,Y_{\ell},X\} adapted to the ascending central series; \mathfrak{z}=\mathbb{R}Z and \mathfrak{v}=\mathrm{span}_{\mathbb{R}}\{Y_{1},\dots,Y_{\ell}\} . • \pi is an irreducible SI/Z representation with central character \pi(\exp(zZ))=e^{2\pi i\,z\,\lambda(Z)}\,\mathrm{Id} for some \lambda\in\mathfrak{z}^{\ast}\setminus\{0\} . • f\in H_{\pi}^{\infty}\setminus\{0\} is a fixed smooth vector. Because N is a connected and simply connected nilpotent Lie group, \exp:\mathfrak{n}\to N is a diffeomorphism. Every g\in N admits a unique factorization g=\exp(W)\exp(tX)\qquad(W\in\mathfrak{z}\oplus\mathfrak{v},\ t\in\mathbb{R}). Set \mathcal{V}(W,t):=\langle f,\pi(\exp(W)\exp(tX))f\rangle,\qquad V(W,\xi):=\int_{\mathbb{R}}\mathcal{V}(W,t)e^{-2\pi it\xi}\,dt. 3.1 Setup and reduction to noncentral variables Let N be a connected and simply connected step-two nilpotent Lie group with one-dimensional center Z(N) and Lie algebra \mathfrak{n} . Fix a basis adapted to the central series \{B_{1}=Z,\ B_{2}=Y_{1},\dots,B_{\ell+1}=Y_{\ell},\ B_{\ell+2}=X\}, meaning \mathfrak{n}(k)=\mathrm{span}_{\mathbb{R}}\{B_{1},\dots,B_{k}\} is an ideal for each k [ 2 ] . Set N_{0}:=\exp(\mathbb{R}Z)\,\exp(\mathbb{R}Y_{1})\cdots\exp(\mathbb{R}Y_{\ell}),\qquad\mathfrak{v}:=\mathrm{span}_{\mathbb{R}}\{Y_{1},\dots,Y_{\ell}\}. Then the map (W,t)\mapsto\exp(W)\exp(tX) is a global diffeomorphism from (\mathfrak{z}\oplus\mathfrak{v})\times\mathbb{R} onto N . Matrix coefficients and partial Fourier transform. For f\in H_{\pi}^{\infty} (smooth for \pi ), define \mathcal{V}(W,t):=\langle f,\ \pi(\exp(W)\exp(tX))\,f\rangle_{H_{\pi}},\qquad(W\in\mathfrak{v},\ t\in\mathbb{R}), (1) and its partial Fourier transform in t , (F_{X}\mathcal{V})(W,\xi):=\int_{\mathbb{R}}\mathcal{V}(W,t)\,e^{-2\pi it\xi}\,dt. (2) In the two-step case one has the standard regularity (see [ 8 , 2 ] ) identifying smooth vectors and giving Schwartz decay of matrix coefficients in the noncentral variables. Proposition 3.1 (Schwartz regularity in noncentral variables) . Assume \mathfrak{n} is two-step. If f\in H_{\pi}^{\infty} , then \mathcal{V}\in\mathcal{S}(\mathfrak{v}\times\mathbb{R})\quad\text{and}\quad(F_{X}\mathcal{V})(\cdot,\xi)\in\mathcal{S}(\mathfrak{v})\ \text{ for each }\xi\in\mathbb{R}. 3.2 Derivation of the key identity Fix 0\neq Y\in\mathfrak{v} and suppose there exist distinct x_{1}<\cdots<x_{m} and nonzero c_{k}\in\mathbb{C} such that \sum_{k=1}^{m}c_{k}\,\pi(\exp(x_{k}X))f=\pi(\exp Y)f. (3) Testing ( 3 ) against \langle f,\pi(g)f\rangle for arbitrary g\in N yields \sum_{k=1}^{m}c_{k}\,\langle f,\pi(\exp(-x_{k}X)g)f\rangle_{H_{\pi}}=\langle f,\pi(\exp(-Y)g)f\rangle_{H_{\pi}}. (4) Write g=\exp(W)\exp(tX) with W\in\mathfrak{z}\oplus\mathfrak{v} , t\in\mathbb{R} . In a two–step algebra one has e^{-x\mathrm{ad}X}W=W-x[X,W],\qquad\log\!\big(\exp(-Y)\exp(W)\big)=W-Y-\tfrac{1}{2}[Y,W], (5) with [X,W],[Y,W]\in\mathfrak{z} . Using ( 5 ), the left side of ( 4 ) becomes \langle f,\pi(\exp(W-x_{k}[X,W])\exp((t-x_{k})X))f\rangle. Step 3: Central phases via the character (signs with conjugate linearity). By Schur’s lemma, \pi(\exp Z)=e^{2\pi i\langle\lambda,Z\rangle}\,\mathrm{Id} for Z\in\mathfrak{z} . Because the inner product is conjugate-linear in the second slot, central scalars that act on the second slot exit the inner product conjugated . Hence \displaystyle\mathcal{V}(W-x_{k}[X,W],t-x_{k}) \displaystyle=e^{\,2\pi i\,x_{k}\langle\lambda,[X,W]\rangle}\,\mathcal{V}(W,t-x_{k}), \displaystyle\langle f,\pi(\exp(-Y)\exp(W)\exp(tX))f\rangle \displaystyle=e^{\,2\pi i\,\tfrac{1}{2}\langle\lambda,[Y,W]\rangle}\,\mathcal{V}(W-Y,t). Step 4: Partial Fourier transform in t . Let V(W,\xi):=\int_{\mathbb{R}}\mathcal{V}(W,t)\,e^{-2\pi it\xi}\,dt\qquad(\xi\in\mathbb{R}). Multiply ( 4 ) by e^{-2\pi it\xi} and integrate in t . On each left–hand term change variables t\mapsto t-x_{k} . Using the formulas above gives \sum_{k=1}^{m}c_{k}\,e^{-2\pi ix_{k}\xi}\,e^{\,2\pi ix_{k}\langle\lambda,[X,W]\rangle}\,V(W,\xi)=e^{\,2\pi i\tfrac{1}{2}\langle\lambda,[Y,W]\rangle}\,V(W-Y,\xi). Equivalently, V(W-Y,\xi)=e^{-2\pi i\tfrac{1}{2}\langle\lambda,[Y,W]\rangle}\,\Big(\sum_{k=1}^{m}c_{k}\,e^{-2\pi ix_{k}(\xi+\langle\lambda,[X,W]\rangle)}\Big)\,V(W,\xi). (6) Introduce the shorthand P(W,\xi):=e^{-2\pi i\tfrac{1}{2}\langle\lambda,[Y,W]\rangle}\,\sum_{k=1}^{m}c_{k}\,e^{-2\pi ix_{k}(\xi+\langle\lambda,[X,W]\rangle)}. (7) Then the key identity takes the compact form V(W-Y,\xi)=P(W,\xi)\,V(W,\xi). (8) Note that only the central commutators [X,W] and [Y,W] enter P(W,\xi) , contributing unimodular phases determined by the