The discrete wave equation with applications to scattering theory and quantum chaos
Carsten Peterson · 2025
Read The discrete wave equation with applications to scattering theory and quantum chaos by Carsten Peterson (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-2512-03015.txt.
Introduction
Waves on a line of points follow rules worth knowing well. Carsten Peterson builds up the basic theory of the discrete wave equation in this 2025 work. His aim is to support study of waves on affine buildings, a newer topic in math.
Peterson works on the integers, the whole numbers in a line. From there he explains known results about waves on regular graphs, where each point has the same number of links. Clear basics help others tackle scattering and quantum chaos. The work gives a firm base for more advanced wave studies. Graph wave results then rest on a firm base built step by step from the line case.
Catalog record
| CREATOR | Carsten Peterson |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 136 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2512-03015
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
The discrete wave equation with applications to scattering theory and quantum chaos Carsten Peterson Abstract . With a view towards studying the multitemporal wave equation on affine buildings recently introduced by Anker–Rémy–Trojan [ 4 ] , we systematically develop the basic properties of the discrete wave equation on ℤ \mathbb{Z} and use this to explain existing results about the wave equation on regular graphs. Furthermore, we explicitly compute the incoming and outgoing translation representations and the scattering operator, in the sense of Lax–Phillips, for regular and biregular trees. Finally, we use the wave equation on biregular graphs to extend a result of Brooks–Lindenstrauss about delocalization of eigenfunctions on regular graphs to the setting of biregular graphs. 1. Introduction The wave equation (say, on \mathbb{R}^{n} ) defined by \frac{\partial^{2}}{\partial t^{2}}u(x,t)=\Delta_{x}u(x,t) , is one of the most classical and well-studied partial differential equations in mathematics and physics. It affords many remarkable properties such as existence and uniqueness of solutions given Cauchy data, finite speed of propagation/Huygens’ principle, d’Alembert’s solution (and more generally solutions expressible as superpositions of plane waves), and conservation of energy. From an algebraic perspective, many of these geometric and analytic properties may be viewed as a consequence of the fact that the wave equation involves applying “the same” (at least when n=1 ) differential operator in space and in time, and that this operator is the simplest non-trivial operator invariant under x\mapsto-x . This perspective shall play a distinguished role in the sequel. In this paper we begin by studying the discrete wave equation on \mathbb{Z} , i.e. functions u:\mathbb{Z}\times\mathbb{Z}\to\mathbb{C} satisfying (1) \displaystyle u(n,t+1)+u(n,t-1)=u(n+1,t)+u(n-1,t). If one were to subtract the term 2u(n,t) from both sides of ( 1 ), one would immediately recognize that ( 1 ) simply involves setting the discrete second derivative in the second (i.e. “time”) variable equal to the discrete second derivative in the first (i.e. “space”) variable. We shall at times refer to this as the flat wave equation to contrast it with the wave equation on trees and graphs to be considered later on. While ( 1 ) is certainly not new, we have been unable to find a systematic treatment in the literature. Note that, if we were to convert a function on \mathbb{Z}\times\mathbb{Z} into a formal bi-infinite Laurent series in x and y , then the LHS of ( 1 ) amounts to multiplication by y+y^{-1} , and the RHS amounts to multiplication by x+x^{-1} . As we shall show, most of the exceptional properties of ( 1 ) are a result of algebraic properties of the Laurent polynomial ring \mathbb{C}[x,x^{-1}] (in fact, many of the remarkable properties of the wave equation on \mathbb{R}^{n} are a result of algebraic properties of \mathbb{C}[x] ; see Section 2.7 ). Let W=S_{2} be the symmetric group on two elements whose non-trivial element w_{0} acts on \mathbb{C}[x,x^{-1}] via x\mapsto x^{-1} . Then \mathbb{C}[x,x^{-1}] is a free module of rank 2 over \mathbb{C}[x,x^{-1}]^{W}=\mathbb{C}[x+x^{-1}] (Proposition 2.1 ). Choosing appropriate initial conditions to guarantee existence and uniqueness of solutions ultimately amounts to choosing a basis for this module. A very natural, and in many ways best possible, basis is given by h_{1}=1,h_{2}=\frac{x-x^{-1}}{2} . Furthermore, \mathbb{C}[x,x^{-1}] carries a natural non-degenerate \mathbb{C}[x+x^{-1}] -bilinear pairing (\cdot,\cdot) taking values in \mathbb{C}[x+x^{-1}] (Proposition 2.2 ). This allows us to find a dual basis for any basis; the dual basis to h_{1},h_{2} is m_{1}=\frac{x-x^{-1}}{2},m_{2}=1 . The relationship between initial conditions and the corresponding fundamental solutions is essentially the relationship between a basis and the corresponding dual basis (Theorem 2.4 ). The d’Alembert presentation of solutions to the wave equation is essentially a consequence of the definition of (\cdot,\cdot) (Corollary 2.6 ), finite speed of propagation is immediate from the explicit expression of the solutions (Corollary 2.5 ), and conservation of energy amounts to diagonalizing the matrix [(m_{j}h_{j},x)] (Proposition 2.17 ). Furthermore, as we discuss in Section 2.5 , studying “fundamental solutions” to ( 1 ) essentially amounts to studying Chebyshev polynomials, or more generally Bernstein-Szegö polynomials. Orthogonality properties of such polynomials can also be understood using this framework. See Theorem 2.9 . When considering the wave equation on the hyperbolic plane \mathbb{H} (or on hyperbolic surfaces), often one studies the shifted wave equation (also referred to as the automorphic wave equation in the context of hyperbolic surfaces) defined as solutions u:\mathbb{H}\times\mathbb{R}\to\mathbb{C} to the equation (2) \displaystyle\frac{\partial^{2}}{\partial t^{2}}u(x,t)=\Big(\Delta_{x}-\frac{1}{4}\Big)u(x,t). See, for instance, [ 26 , 19 , 21 , 2 , 5 ] . The appearance of \Delta-\frac{1}{4} on the RHS may be explained as follows. There is a natural algebra isomorphism \Gamma (a special case of the Harish-Chandra isomorphism) from \textnormal{SL}(2,\mathbb{R}) -invariant differential operators on \mathbb{H} to polynomials on \mathbb{C} invariant under x\mapsto-x ; the latter is simply \mathbb{C}[x^{2}] . We have that \Gamma(\Delta-\frac{1}{4})=x^{2} , which is the symbol of the operator \frac{\partial^{2}}{\partial x^{2}} which appears on the RHS of the wave equation on \mathbb{R} . The wave equation ( 2 ) has nicer properties than the usual wave equation on \mathbb{H} , partially because it interfaces much better with the representation theory of \textnormal{SL}(2,\mathbb{R}) . One may ultimately study ( 2 ) by a two-step process: first by studying the wave equation on \mathbb{R} , and then studying the nature of the map \Gamma (the latter of which can be studied via spherical harmonic analysis on \mathbb{H} ). Let \mathcal{T} be the infinite (q+1) -regular tree. The discrete analogue of ( 2 ) is (3) \displaystyle u(v,t+1)+u(v,t-1)=\Big(\frac{A}{\sqrt{q}}\Big)u(v,t) where u:\mathcal{T}\times\mathbb{Z}\to\mathbb{C} , and A is the adjacency operator on \mathcal{T} (we shall only consider functions on the vertices of \mathcal{T} ). Note that ( 3 ) was independently and roughly simultaneously introduced by [ 8 ] as the p -adic wave equation (see also [ 7 , 6 ] ), and by [ 1 ] as the shifted wave equation . In fact, we also found the same equation considered in the much earlier work [ 28 ] . There is a natural map Sat (a special case of the Satake isomorphism) from \textnormal{Aut}(\mathcal{T}) -invariant operators on \mathcal{T} (also known as the spherical Hecke algebra) to \mathbb{C}[x,x^{-1}]^{W} . We have that \textnormal{Sat}(\frac{A}{\sqrt{q}})=x+x^{-1} , which is essentially the “symbol” of the operator appearing on the RHS of ( 3 ). After discussing the relevant tools from spherical harmonic analysis on \mathcal{T} in Section 3.1 , we quickly and systematically recover in Section 3.2 the main results of [ 8 ] and [ 1 ] concerning ( 3 ). We remark that many ideas closely related to the wave equation on the regular tree have implicitly appeared for a long time in analytic number theory in terms of Chebyshev polynomials of the normalized Hecke operator; see for instance [ 31 , 10 , 16 , 29 ] . In case q is a power of a prime, we may view \mathcal{T} as the Bruhat–Tits tree associated to \textnormal{SL}(2,F) , with F a non-archimedean local field with residue field of order q , and we may view the adjacency operator A as the Hecke operator, which plays an important role in p -adic representation theory and the theory of automorphic forms. Lax–Phillips established a general framework for studying wave-like equations, often referred to as the Lax–Phillips scattering theory .