St. Petersburg Department of Steklov Mathematical Institute of Russian Academy of Sciences,
27 Fontanka, St. Petersburg 191023, Russia · 2025
Read St. Petersburg Department of Steklov Mathematical Institute of Russian Academy of Sciences, by 27 Fontanka, St. Petersburg 191023, Russia (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-2504-12818.txt.
Introduction
Ivanov studies renormalization in a short 2025 maths paper. Renormalization is a procedure for removing infinities, called singularities, that appear in certain calculations. The author examines one special model where the problem can be attacked in two distinct ways.
One route uses formal series and counting-style techniques. The other takes a limit inside a standard multi-dimensional integral, with the singular parts stripped out first. A special bending of the spectrum serves as the regulator that keeps values finite. Ivanov stresses a striking contrast: sums that look alike in one view take a different nature in the other. That contrast leads to results he finds notable, and the note compares the two views in detail.
Catalog record
| CREATOR | 27 Fontanka, St. Petersburg 191023, Russia |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 53 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2504-12818
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
St. Petersburg Department of Steklov Mathematical Institute of Russian Academy of Sciences, 27 Fontanka, St. Petersburg 191023, Russia Abstract An etude on a renormalization Aleksandr V. Ivanov E-mail: [email protected] On the 85th anniversary of PDMI RAS A.A.Ivanov (1936–1996) Abstract. In this paper, we study renormalization, that is, the procedure for eliminating singularities, for a special model using both combinatorial techniques in the framework of working with formal series, and using a limit transition in a standard multidimensional integral, taking into account the removal of the singular components. Special attention is paid to the comparative analysis of the two views on the problem. It is remarkably that the divergences, which have the same form in one approach, acquire a different nature in another approach and lead to interesting consequences. A special deformation of the spectrum is used as regularization. Keywords and phrases: renormalization, regularization, spectrum deformation, Gaussian integral, diagrammatic technique, formal series, multidimensional integral. 1 Introduction Regularization and renormalization play an important role in modern theoretical physics and are used in a large number of quantum field models [ 1 , 2 , 3 , 4 , 5 , 6 ] . However, in mathematical physics and pure mathematics, such concepts are rare or have a slightly different meaning, see [ 7 , 8 ] . First of all, this is due to the fact that the mathematical formalism devoted to a functional integration, see [ 9 , 10 , 11 ] , is currently insufficiently developed. It is for this reason that the basic quantum field models have to be studied using the perturbative method, for which it is quite sufficient to understand the functional integral in a very stripped-down form. From the point of view of the perturbative approach, which is associated with the analysis of formal (less often asymptotic) series, the concepts of regularization and renormalization are led by the presence of non-integrable densities and singular functionals in the coefficients of the series under study. This approach in this article will be called ”physical”. However, within the framework of another approach, which will be called ”mathematical”, in those models in which a strict definition of the functional integral can be used, formal series no longer appear, and consequently, the reasons for the introduction of the regularization and the application of the renormalization become different. Note that the words ”physical” and ”mathematical” are used only to distinguish, both approaches have a neat mathematical form. In this paper, we analyze a special type of model that allows for both the physical description and a formulation using the functional integral in the sense of limit transition in a finite-dimensional case, see [ 12 , 13 , 14 ] . The latter theme is quite popular and is used for various models [ 15 , 16 , 19 , 17 , 18 ] . This formulation of the problem allows us to trace all the subtleties of both approaches, make a comparative analysis, and identify the positive and negative sides. Remarkably, the divergences that arise when using the perturbative approach due to the appearance of non-integrable densities take on a completely different meaning in the second approach: they manifest themselves either by the presence of a rapidly oscillating phase, or by the zeros of the entire functional. At the same time, the regularization and the renormalization are performed in the same way in both approaches. The purpose of this paper is to demonstrate the application of mathematical formalism to a process that often occurs in theoretical physics, as well as an attempt to draw the attention of researchers working in related mathematical fields to an existing interesting problem. The mathematical analysis of the model presented in the article can be considered a brief outline, demonstrating in what form the renormalization procedure would like to generalize to a wider class of models. The work has the following structure. Section 2 studies the case using a special type of functional integral. Necessary definitions and proofs are provided. Concepts such as the regularization, the renormalization, and the limit transition are discussed. Section 3 examines an approach based on the use of formal series. Concepts such as a diagrammatic technique, as well as the regularization and the renormalization are discussed. The section also contains a comparison of approaches. Section 4 contains some concluding remarks. 2 Mathematical approach Definition 1 . Let V denote a set of sequences a=\{a_{j}\}_{j=1}^{+\infty} , the elements a_{j} of which belong to \mathbb{R} . Let also n\in\mathbb{N} . The symbol p_{n} denotes a projector from V to V acting on an arbitrary a\in V according to the rule: \big(p_{n}(a)\big)_{j}=a_{j}\,\,\,\mbox{for}\,\,\,j\in\{1,\ldots,n\},\,\,\,\big(p_{n}(a)\big)_{j}=0\,\,\,\mbox{for}\,\,\,j>n. (1) Definition 2 . Let k\in\mathbb{N} . With the symbol \mathcal{B}_{k} , we notate a subset of V , all elements \beta\in\mathcal{B}_{k} of which satisfy two conditions: 1. \beta_{j}>0 for all j\in\mathbb{N} ; 2. b_{k}(\beta)=\sum_{j=1}^{+\infty}\beta_{j}^{-k}<+\infty . In addition, we use the symbol \mu(\beta) to notate the minimum element, that is \min_{j>0}(\beta_{j}) . Remark 1 . If \beta\in\mathcal{B}_{k} for some k\in\mathbb{N} , then \beta\in\mathcal{B}_{j} for all k\leqslant j\in\mathbb{N} . Also note that, if necessary, the elements of the sequence \beta can be ordered in ascending order, since the only point of accumulation is infinity. Definition 3 . Let n,k\in\mathbb{N} , s\in\mathbb{R} , and also \beta\in\mathcal{B}_{k} . Let us introduce the function \Phi_{n}(s,\beta)=\prod_{j=1}^{n}\int_{\mathbb{R}}\frac{\mathrm{d}a_{j}}{\sqrt{\pi/\beta_{j}}}\,e^{-\beta_{j}^{\phantom{1}}a_{j}^{2}+isa_{j}^{2}}. (2) Remark 2 . Note that the integral representation ( 2 ) is set correctly for all s\in\mathbb{C} , such that \mathrm{Im}(s)>-\mu(\beta) . In this paper, the assumption \mathrm{Im}(s)=0 is not a significant limitation, since it can be achieved by changing the normalization of the integral and redefining the elements \beta_{j}\to\tilde{\beta}_{j}=\beta_{j}-\mathrm{Im}(s) . Example 1 . In mathematical physics, the set \beta is usually called a spectrum and is directly related to differential operators. Consider the following example. Let \mathbb{N}\ni n>1 and \mathbf{B}=\{x\in\mathbb{R}^{n}:|x|\leqslant 1\} is the n -dimensional ball. In the domain \mathbf{B} , we can consider a spectral problem with zero Dirichlet boundary condition -\sum_{i=1}^{n}\partial_{x_{i}}^{2}u(x)=\lambda u(x),\,\,\,u(x)\big|_{|x|=1}=0. It is known, see paragraph 4 of chapter 5 in [ 20 ] , that this problem has a countable spectrum \beta=\{\beta_{j}\}_{j=1}^{+\infty} and the corresponding set of eigenfunctions \{\phi_{j}(x)\in L^{2}(\mathbf{B})\}_{j=1}^{+\infty} . In this case, the set \beta has a single accumulation point at infinity, and all its elements are strictly positive. The Green’s function for the mentioned operator in this case has the form G(x,y)=\sum_{j=1}^{+\infty}\phi_{j}(x)\beta^{-1}_{j}\phi_{j}(y), where x,y\in\mathbf{B} and x\neq y . Let us pay attention to the behavior of the Green’s function near the diagonal x\sim y , see [ 21 , 22 ] , which in the main order, when decomposed in x-y , is proportional to the function \ln(|x-y|)\,\,\,\mbox{for}\,\,\,n=2\,\,\,\mbox{and}\,\,\,|x-y|^{2-n}\,\,\,\mbox{for}\,\,\,n>2. It is clear that in this case b_{1}(\beta)=+\infty for all n>1 , since the value of G(x,y) diverges at y\to x inside the domain \mathbf{B} . The convergence of the remaining numbers b_{k}(\beta) , where k>1 , is conveniently analyzed using the representation b_{k}(\beta)=\int_{\mathbf{B}}\mathrm{d}^{n}x_{1}\ldots\int_{\mathbf{B}}\mathrm{d}^{n}x_{k}\,G(x_{1},x_{2})\cdot\ldots\cdot G(x_{k-1},x_{k}). In particular, if n=2,3, then in this case \beta\in\mathcal{B}_{2} occurs due to the convergence of the last integrals. Lemma 1 . Let the assumptions of Definition 3 be true. Then for function ( 2 ) the following representation is valid \Phi_{n}(s,\beta)=\Bigg(\prod_{j=1}^{+\infty}\int_{\mathbb{R}}\frac{\mathrm{d}a_{j}}{\sqrt{\pi/\beta_{j}}}\Bigg)e^{-S_{0}[a,\beta]+isS_{1}[p_{n}(a)]}, (3) where auxiliary objects are defined by the equalities S_{0}[a,\beta]=\sum_{j=1}^{+\infty}\beta_{j}^{\phantom{1}}a_{j}^{2},\,\,\,S_{1}[a]=\sum_{j=1}^{+\infty}a_{j}^{2}. (4) Proof. Note that the object from ( 3