Stochastic motions of the two-dimensional many-body delta-Bose gas, III: Path integrals

Yu-Ting Chen · 2025

Read Stochastic motions of the two-dimensional many-body delta-Bose gas, III: Path integrals by Yu-Ting Chen (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-2505-03006.txt.

Introduction

Chen studies random motions of many bosons that pull on each other at a single point, in a paper from twenty twenty-five. The work is the third part in a series on this gas in two dimensions. The gas model is used to test hard ideas in quantum physics. Chen aims to link the random motions to path-sum formulas.

He uses motions built in the prior part of the series as the base paths. He shows the path-sum formulas hold for any count of three or more bosons. The weights on the paths take a new form here. They come from exact solutions found in an earlier paper in the series. For two bosons, the formula is a small tweak of a known one.

Catalog record

CREATORYu-Ting Chen
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE122 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2505-03006

Browse the full library · About the Tower of Babel

Full text

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

Stochastic motions of the two-dimensional many-body delta-Bose gas, III: Path integrals Support from an NSERC Discovery grant is gratefully acknowledged. Yu-Ting Chen Abstract This paper is the third in a series devoted to constructing stochastic motions for the two-dimensional N N -body delta-Bose gas for all integers N ≥ 3 N\geq 3 and establishing the associated Feynman–Kac-type formulas. The main results here prove the Feynman–Kac-type formulas by using the stochastic many- δ \delta motions from [ 7 ] as the underlying diffusions. The associated multiplicative functionals show a new form and are derived from the analytic solutions of the two-dimensional N N -body delta-Bose gas obtained in [ 4 ] . For completeness, the main theorem includes the formula for N = 2 N=2 , which is a minor modification of the Feynman–Kac-type formula proven in [ 5 ] for the relative motions. Keywords: Delta-Bose gas; Schrödinger operators; diagrammatic expansions; path integrals; Bessel processes. Mathematics Subject Classification (2020): 60J55, 60J65, 60H30, 81S40. Contents 1 Introduction This paper is the third in a series devoted to constructing stochastic motions for the two-dimensional N -body delta-Bose gas for all integers N\geq 3 and establishing the associated Feynman–Kac-type formulas. We will prove the formulas in this paper by using a {\mathbb{C}}^{\it N} -valued strong Markov process \{\mathscr{Z}_{t}\}=\{Z^{j}_{t}\}_{1\leq j\leq N} , called a stochastic many- \boldsymbol{\delta} motion . It has continuous paths before its lifetime T_{\partial} and is defined under the probability measure {\mathbb{P}}^{{\boldsymbol{\beta}},\boldsymbol{w}} constructed in [ 7 ] . Here, {\boldsymbol{\beta}},\boldsymbol{w}\in(0,\infty)^{\mathcal{E}_{N}} are parameters tuning strengths of interactions of particles in different manners and play the roles of coupling constants and weights, respectively, where \displaystyle\mathcal{E}_{N}\,{\stackrel{{\scriptstyle\rm def}}{{=}}}\,\{\mathbf{j}=(j\prime,j)\in\mathbb{N}^{2};1\leq j<j\prime\leq N\}. (1.1) The main properties of this stochastic many- \delta motion can be found in [ 7 , Theorem 3.1, Proposition 3.11 and Proposition 3.15] , and some of them will be recalled later. Note that we only consider w_{\mathbf{j}}>0 for all \mathbf{j}\in\mathcal{E}_{N} in this paper although this condition is not required in [ 7 ] . Consequently, the notations here differ slightly; see [ 7 , Remark 3.2] . Also, [ 7 ] includes a description of this stochastic many- \delta motion by using an SDE with singular drift explicitly expressible in ({\boldsymbol{\beta}},\boldsymbol{w}) , but the SDE will be used here only in Remark 2.4 . The primary applications of the above stochastic many- \delta motion in this paper are based on its interpretation as a family of N independent two-dimensional Brownian motions conditioned for contact interactions. These interactions are realized in the form of Z_{t}^{j\prime}-Z_{t}^{j}=0 for \mathbf{j}=(j\prime,j)\in\mathcal{E}_{N} to the degree of nonzero local times \{L^{\mathbf{j}}_{t}\} and satisfy the no-triple-contacts (NTC) that at most two particles can engage in contact interactions at a time. Moreover, the sample paths of the stochastic many- \delta motion can be subdivided by an increasing sequence of stopping times \{T^{m}_{0}\}_{m\in\mathbb{N}} , called the contact-creation times , according to the different pairs of particles that can engage in contact interactions. Specifically, in [T_{0}^{m},T_{0}^{m+1}) for any m\geq 0 ( T_{0}^{0}\equiv 0 ), no contact interactions occur in particles with different states at time T_{0}^{m} , and whenever m\geq 1 , contact interactions occur in a unique pair of particles, labelled by a random index \boldsymbol{J}_{m}=(J_{m}\prime,J_{m})\in\mathcal{E}_{N} . These characteristics make possible the construction of the stochastic many- \delta motion by locally transforming a stochastic one- \boldsymbol{\delta} motion from [ 5 , 6 ] with Girsanov’s theorem over [T_{0}^{m},T_{0}^{m+1}) . Nonetheless, the continuous extension to T_{0}^{m+1} concerns constructing and studying the exit measure of a supermartingale. New properties are elicited from the extremal behavior of the stochastic one- \delta motion for changing measures, including especially the NTC at T_{0}^{m+1} to restart the construction over the next interval [T_{0}^{m+1},T_{0}^{m+2}) . 1.1 Analytic path integrals The Feynman–Kac-type formulas in this paper represent limiting semigroups of the following approximate Hamiltonians for the two-dimensional N -body delta-Bose gas: \displaystyle-\frac{1}{2}\sum_{j=1}^{N}\Delta_{z^{j}}-\sum_{\mathbf{j}=(j\prime,j)\in\mathcal{E}_{N}}\left(\frac{2\pi}{\log{\varepsilon}^{-1}}+\frac{2\pi\lambda_{\mathbf{j}}}{\log^{2}{\varepsilon}^{-1}}\right){\varepsilon}^{-2}\phi_{\mathbf{j}}({\varepsilon}^{-1}(z^{j\prime}-z^{j})),\quad z^{j}\in\mathbb{C}, (1.2) where \lambda_{\mathbf{j}} and \phi_{\mathbf{j}} can use any given real constants and probability densities in {\mathscr{C}}_{c}({\mathbb{R}}^{2}) , respectively. Note that these approximations show a renormalization generalizing the two-particle case introduced in [ 1 ] ; see [ 6 , Section 1] for a discussion of this setting. Specifically, we will work with the analytic solutions Q^{{\boldsymbol{\beta}};N}_{0;t}f(z_{0}) obtained as the limits of the following Feynman–Kac semigroups representing the approximate Hamiltonians in ( 1.2 ): \displaystyle\quad\;\lim_{{\varepsilon}\to 0}{\mathbb{E}}^{(0)}_{z_{0}}\Biggl[\exp\Biggl\{\sum_{\mathbf{j}=(j\prime,j)\in\mathcal{E}_{N}}\int_{0}^{t}\left(\frac{2\pi}{\log{\varepsilon}^{-1}}+\frac{2\pi\lambda_{\mathbf{j}}}{\log^{2}{\varepsilon}^{-1}}\right){\varepsilon}^{-2}\phi_{\mathbf{j}}({\varepsilon}^{-1}(Z^{j\prime}_{r}-Z^{j}_{r})){\mathrm{d}}r\Biggl\}f(\mathscr{Z}_{t})\Biggr] \displaystyle=Q^{{\boldsymbol{\beta}};N}_{0;t}f(z_{0}),\quad\forall\;f\in\mathscr{B}_{b}({\mathbb{C}}^{\it N}). (1.3) Here, z_{0}=(z_{0}^{1},\cdots,z_{0}^{N})\in{\mathbb{C}}^{\it N} satisfies z_{0}^{j\prime}\neq z_{0}^{j} for all j\prime\neq j , \{\mathscr{Z}_{t}\}=\{Z^{j}_{t}\}_{1\leq j\leq N} under {\mathbb{P}}_{z_{0}}^{(0)} is a 2N -dimensional standard Brownian motion with initial condition z_{0} , and \displaystyle Q^{{\boldsymbol{\beta}};N}_{0;t}f(z_{0})\,{\stackrel{{\scriptstyle\rm def}}{{=}}}\,{\mathbb{E}}^{(0)}_{z_{0}}\bigl[f(\mathscr{Z}_{t})\bigr]+\sum_{m=1}^{\infty}\sum_{\begin{subarray}{c}{\mathbf{i}}_{1},\ldots,{\mathbf{i}}_{m}\in\mathcal{E}_{N}\\ {\mathbf{i}}_{1}\neq\cdots\neq{\mathbf{i}}_{m}\end{subarray}}\int_{0<s_{1}<\cdots<s_{m}<t}P^{{\boldsymbol{\beta}};{\mathbf{i}}_{1},\ldots,{\mathbf{i}}_{m}}_{s_{1},\ldots,s_{m},t}f(z_{0})\mathrm{d}\boldsymbol{s}_{m}. (1.4) The definition in ( 1.4 ) uses {\mathrm{d}}\boldsymbol{s}_{m}\,{\stackrel{{\scriptstyle\rm def}}{{=}}}\,{\mathrm{d}}s_{1}\cdots{\mathrm{d}}s_{m} and P^{{\boldsymbol{\beta}};{\mathbf{i}}_{1},\ldots,{\mathbf{i}}_{m}}_{s_{1},\ldots,s_{m},t}f(z_{0}) from ( 2.3 ) as an iterated space-time Riemann integral of products of copies of the following functions: \displaystyle P_{s^{\prime}-s}(z,z^{\prime})\,=P_{s^{\prime}-s}(z^{\prime}-z)\;{\stackrel{{\scriptstyle\rm def}}{{=}}}\,\frac{1}{2\pi(s^{\prime}-s)}\exp\biggl(-\frac{|z^{\prime}-z|^{2}}{2(s^{\prime}-s)}\biggr), (1.5) \displaystyle{\mathfrak{s}}^{\widetilde{\beta}}(\tau)\;{\stackrel{{\scriptstyle\rm def}}{{=}}}\,4\pi\int_{0}^{\infty}\frac{\widetilde{\beta}^{u}\tau^{u-1}}{\Gamma(u)}{\mathrm{d}}u. (1.6) Also, the set {\boldsymbol{\beta}} of coupling constants has components \beta_{\mathbf{j}} determined by the following equations: \displaystyle\frac{\log\beta_{\mathbf{j}}}{2}=\log 2+\lambda_{\mathbf{j}}-\gamma_{\mathsf{E}\mathsf{M}}-\int_{\mathbb{C}}\int_{\mathbb{C}}\phi_{\mathbf{j}}(z)\phi_{\mathbf{j}}(z^{\prime})\log|z-z^{\prime}|{\mathrm{d}}z{\mathrm{d}}z^{\prime}, where \gamma_{\mathsf{E}\mathsf{M}} is the Euler–Mascheroni constant. Note that ( 1.1 ) can be justified by the proof of [ 4 ] mutatis mutandis , although [ 4 ] only considers the homogeneous case of \lambda_{\mathbf{j}}\equiv\lambda and \phi_{\mathbf{j}}\equiv\phi ; see [ 9 ] for the analytic solutions of the homogeneous case under f\in L^{2} . 0 s_{1} \tau_{1} s_{2} \tau_{2} s_{3} \tau_{3} s_{4} \tau_{4} s_{5} \tau_{5} t \bullet \bullet \bullet \bullet z_{0}^{1} z_{0}^{2} z_{0}^{3} z_{0}^{4} \bullet \bullet \bullet \bullet \bullet \bullet \bullet \bullet \bullet \bullet \bullet \bullet \bullet \bullet \bullet \bullet \bullet \bullet \bullet \bullet \bullet \bullet \bullet \bullet \bullet \bullet \bullet \bullet \bullet \bullet \bullet \bullet \bullet \bullet \bullet \bullet \bullet \bullet \bullet \bullet \bullet \bullet \bullet \bullet Figure 1.1: The figure illustrates the graphical representation of P^{{\boldsymbol{\beta}};\mathbf{i}_{1},\mathbf{i}_{2},\mathbf{i}_{3},\mathbf{i}_{4},\mathbf{i}_{5}}_{s_{1},s_{2},s_{3},s_{4},s_{5},t}f(z_{0}) defined in [ 4 ] , with N=4 , \mathbf{i}_{1}=(2,1) , \mathbf{i}_{2}=(3,2) , \mathbf{i}_{3}=(2,1) , \mathbf{i}_{4}=(4,3) and \mathbf{i}_{5}=(2,1) . Despite the notational complexity in ( 2.3 ), each P^{{\boldsymbol{\beta}};{\mathbf{i}}_{1},\ldots,{\mathbf{i}}_{m}}_{s_{1},\ldots,s_{m},t}f(z_{0}) mentioned above has been known for simple interpretations (e.g. [ 3 , 4 , 9 ] ). For example, the interpretation in [ 4 , Section 4] uses a graphical representation suggesting paths of N particles with pairwise contact interactions. Specifically, with respect to the space-time integration in ( 2.3 ), the graphical representation from [ 4 ] views the integrand of each kernel, excluding f , as a product of edge weights of two types such that every edge weight defines a line segment in space-time: • The edge weight of the first type takes the form of a heat kernel P_{s^{\prime}-s}(\mbox{cnst}\cdot z,z^{\prime}) and defines a straight line segment from (z,s) to (z^{\prime},s^{\prime}) . This line segment represents the path of one particle, which is free in the sense of not being affected by any contact interaction. • The edge weight of the second type takes the form of a product of {\mathfrak{s}}^{\widetilde{\beta}}(s^{\prime}-s)P_{s^{\prime}-s}(\mbox{cnst}\cdot z,z^{\prime}) and defines a coiled line segment from (z,s) and (z^{\prime},s^{\prime}) . This line segment simultaneously represents the paths of two particles undergoing contact interactions. The overall interpretation of the graphical representation from [ 4 ] thus considers a discrete graph such that the vertices are distinct pairs of the form (\mathring{z},\mathring{s})\in{\mathbb{R}}^{2}\times{\mathbb{R}}_{+} . Here, \mathring{z} is chosen from the space variables in ( 2.3 ) for integration (Remark 2.2 ) or is one of the distinct points z_{0}^{1},\cdots,z_{0}^{N} in the initial condition; \mathring{s}

Read the complete text