Bochkarev’s inequalities in the Anisotropic grand Lorentz Spaces
Nazerke T. Tleukhanova, Makhpal Manarbek · 2025
Read Bochkarev’s inequalities in the Anisotropic grand Lorentz Spaces by Nazerke T. Tleukhanova, Makhpal Manarbek (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-2512-15320.txt.
Introduction
Nazerke T. Tleukhanova and a coauthor extend classic inequalities about Fourier series into a broader family of function spaces. Bochkarev's inequalities, of Hardy-Littlewood type, link the integral behavior of a function to how summable its Fourier coefficients are. The classical results describe trigonometric series in Lorentz spaces with second index above 2.
The new work moves to anisotropic grand Lorentz spaces, where the measure of a function can differ by direction and includes a grand, limiting flavor. Using extrapolation of linear operators, the authors derive Bochkarev type inequalities for these spaces and for arbitrary orthonormal systems, not just trigonometric ones. They also study the summability of Fourier coefficients for functions in the new spaces. The results generalize a classical chapter of harmonic analysis to a much wider setting.
Catalog record
| CREATOR | Nazerke T. Tleukhanova, Makhpal Manarbek |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 45 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2512-15320
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Bochkarev’s inequalities in the Anisotropic grand Lorentz Spaces Nazerke T. Tleukhanova, Makhpal Manarbek Abstract . The main aim of this paper is to obtain Bochkarev-type inequalities for the anisotropic grand Lorentz spaces. In the classical setting, Bochkarev obtained inequalities of the Hardy–Littlewood type, which reveal the connection between the integral properties of functions and the summability of their Fourier coefficients. His results describe the behavior of trigonometric series in the Lorentz spaces L 2 , r L_{2,r} for 2 < r ≤ ∞ 2<r\leq\infty . In this work, we extend these ideas to the framework of anisotropic grand Lorentz spaces. Using an approach based on the extrapolation of linear operators, we derive new Bochkarev-type inequalities that generalize the classical results to the case of anisotropic grand Lorentz spaces and arbitrary orthonormal systems. We investigate the summability properties of the Fourier coefficients of functions from anisotropic grand Lorentz spaces. 1. Introduction One of the principal directions of harmonic analysis is the theory of Fourier series [ 1 , 2 , 3 , 4 ] . Interest in this area is driven by its applications across modern mathematics and the applied sciences, as well as by the presence of many difficult open problems [ 5 , 6 , 7 , 8 , 9 , 10 , 11 , 12 , 13 , 14 , 15 , 16 ] . Parseval’s identity is perhaps the most familiar result of this type [ 17 ] . Let f be an integrable function on [0,1] , and let \{\varphi_{k}\}_{k=1}^{\infty} be an orthonormal system in L_{2}([0,1]) . The Fourier coefficients of f with respect to this system are defined by a_{k}=a_{k}(f):=\int_{0}^{1}f(x)\,\overline{\varphi_{k}(x)}\,dx. When f\in L_{2}([0,1]) , Parseval’s identity asserts that \|f\|_{L_{2}}=\|a\|_{\ell_{2}}:=\left(\sum_{k\in\mathbb{Z}}|a_{k}|^{2}\right)^{1/2}, where a_{k}=\int_{0}^{1}f(x)e^{-2\pi ikx}\,dx are the standard Fourier coefficients associated with the trigonometric orthonormal system. Historically in 1956, Stein [ 18 ] proved the Hardy–Littlewood inequality for Lorentz spaces, which generalizes Paley’s inequality: (1) If 1<p<2 , 0<q\leq\infty , p^{\prime}=\tfrac{p}{p-1} , and f\in L_{p,q}[0,1] , then \left(\sum_{k=1}^{\infty}k^{\tfrac{q}{p^{\prime}}-1}(a_{k}^{*})^{q}\right)^{\tfrac{1}{q}}\leq c\,\|f\|_{L_{p,q}[0,1]}. (2) If 2<p<\infty , 0<q\leq\infty , and \sum_{k=1}^{\infty}k^{\tfrac{q}{p^{\prime}}-1}(a_{k}^{*})^{q}<\infty , then there exists f\in L_{p,q} such that a_{n}(f)=a_{n} , and \|f\|_{L_{p,q}[0,1]}\leq c\left(\sum_{k=1}^{\infty}k^{\tfrac{q}{p^{\prime}}-1}(a_{k}^{*})^{q}\right)^{\tfrac{1}{q}}. Here, for p=q one recovers Paley’s inequality. In the case 1<p<2 , all these inequalities provide necessary conditions for a function to belong to L_{p} or L_{p,q} , while for 2<p<\infty they provide sufficient conditions. The problem of estimating Fourier coefficients of functions in Lorentz spaces L_{2,q} , with 2<q\leq\infty , remained open until the late 20th century. Only in 1997, Bochkarev [ 16 ] established that, unlike in the case of L_{p,r} with 1<p<2 , 1\leq r\leq\infty , in Lorentz spaces L_{2,q} , 2<q\leq\infty , a direct analogue of the Hausdorff–Young–Riesz theorem does not hold. He obtained upper estimates for the Fourier coefficients of functions in L_{2,r} that replace the Hausdorff–Young–Riesz theorem, and proved that for a certain class of multiplicative systems these estimates are sharp. Theorem 1.1 . [ 16 ] Let \{\varphi_{n}\}_{n=1}^{\infty} be an orthonormal system of complex-valued functions on [0,1] satisfying \|\varphi_{n}\|_{\infty}\leq M,\quad n=1,2,\ldots, and let f\in L_{2,q} with 2<q\leq\infty . Then the following inequality holds: \sup_{n\in\mathbb{N}}\frac{1}{n^{1/2}\,(\log(n+1))^{1/2-1/q}}\sum_{m=1}^{n}a_{m}^{*}\leq C\,\|f\|_{L_{2,q}}, where a_{n} are the Fourier coefficients with respect to the system \{\varphi_{n}\}_{n=1}^{\infty} . Further investigations of the behavior of Fourier coefficients of functions belonging to the space L_{2,q} were carried out in the works [ 19 , 20 , 21 ] . In papers [ 22 , 23 ] , refinements of Bochkarev’s inequality were obtained independently of each other. Let \Psi=\{\psi_{n_{1}}(x_{1}),\phi_{n_{2}}(x_{2})\}_{n_{1},n_{2}\in\mathbb{N}},\,\;(x_{1},x_{2})\in[0,1]^{2} , be an orthonormal systems in L_{2}[0,1]^{2} such that for every n\in\mathbb{N} we have |\psi_{n}(x_{1})|\leq M_{1} and |\psi_{n}(x_{2})|\leq M_{2} almost everywhere. Let f\in L_{1}{[0,1]^{2}} , a_{k_{1}k_{2}}=\int_{0}^{1}\int_{0}^{1}f(x_{1},x_{2})\overline{\psi_{k_{1}}(x_{1})}\;\overline{\phi_{k_{2}}(x_{2})}\,dx_{1}dx_{2},\;\;\;\;\;\;\;(k_{1},k_{2})\in\mathbb{N}^{2} the Fourier coefficients of a function with respect to the system \Psi The decreasing rearrangement of f is defined by f^{*}(t)=\inf\{\sigma:\mu\{x\in\Omega:|f(x)|>\sigma\}\leqslant t\}. f^{*_{1},*_{2}}\left(t_{1},t_{2}\right) is obtained from f(x)=f\left(x_{1},x_{2}\right) by applying the nonincreasing rearrangement successively with respect to each variable x_{1} and x_{2} , while keeping the other variable fixed. where rearrangement is successively preformed with respect to each of the variables x_{1} and x_{2} , whereas the other variable is assumed to be fixed (see [ 21 ] ). Let 0<\bar{p}=(p_{1},p_{2})<\infty and 0<\bar{q}=(q_{1},q_{2})\leq\infty . The anisotropic Lorentz space L_{\bar{p},\bar{q}}[0,1]^{2} is defined as the set of all measurable functions f for which the following quantity is finite: \|f\|_{L_{\bar{p},\bar{q}}[0,1]^{2}}=\left(\int_{0}^{1}\left(\int_{0}^{1}\left(t_{1}^{\frac{1}{p_{1}}}t_{2}^{\frac{1}{p_{2}}}f^{*_{1},*_{2}}\left(t_{1},t_{2}\right)\right)^{q_{1}}\frac{dt_{1}}{t_{1}}\right)^{\frac{q_{2}}{q_{1}}}\frac{dt_{2}}{t_{2}}\right)^{\frac{1}{q_{2}}} Here and throughout, expressions of the form \left(\int|\phi(x)|^{q}\frac{dt}{t}\right)^{1/q},\;\left(\sum_{k}|a_{k}|^{q}\frac{1}{k}\right)^{1/q} are understood, in the case q=\infty , as \sup_{x}|\phi(x)|,\;\sup_{k}|a_{k}| respectively. Nursultanov E. D. obtained Hardy–Littlewood–Stein [ 24 ] type inequalities for anisotropic Lorentz spaces. In particular, the following theorems hold: Theorem 1.2 . [ 24 ] Let 0<\bar{q}=(q_{1},q_{2})\leq\infty , \Psi=\{\psi_{n_{1}}(x_{1}),\phi_{n_{2}}(x_{2})\}_{n_{1},n_{2}\in\mathbb{N}} be an orthonormal system of complex-valued functions uniformly bounded in the sense of complete uniformity. If 1<\bar{p}=(p_{1},p_{2})<2 and f\in L_{\bar{p},\bar{q}}[0,1]^{2} , then a(f)\in l^{*}_{\bar{p}^{\prime},\bar{q}} and the inequality \left(\sum_{k_{2}\in\mathbb{N}}\left(\sum_{k_{1}\in\mathbb{N}}\left(k_{2}^{\frac{1}{p^{\prime}_{2}}}k_{1}^{\frac{1}{p^{\prime}_{1}}}a_{k_{1}k_{2}}^{*_{2},*_{1}}\right)^{q_{1}}\frac{1}{k_{1}}\right)^{\frac{q_{2}}{q_{1}}}\frac{1}{k_{2}}\right)^{\frac{1}{q_{2}}}\lesssim\|f\|_{L_{\bar{p},\bar{q}}[0,1]^{2}}, holds. If 2<\bar{p}=(p_{1},p_{2})<\infty и a(f)\in l^{*}_{\bar{p},\bar{q}} , then f\in L_{\bar{p}^{\prime},\bar{q}}[0,1]^{2} and the inequality \|f\|_{L_{\bar{p}^{\prime},\bar{q}}[0,1]^{2}}\lesssim\left(\sum_{k_{2}\in\mathbb{N}}\left(\sum_{k_{1}\in\mathbb{N}}\left(k_{2}^{\frac{1}{p_{2}}}k_{1}^{\frac{1}{p_{1}}}a_{k_{1}k_{2}}^{*_{2},*_{1}}\right)^{q_{1}}\frac{1}{k_{1}}\right)^{\frac{q_{2}}{q_{1}}}\frac{1}{k_{2}}\right)^{\frac{1}{q_{2}}}, holds. is valid, where a(f)=\{a_{k_{1}k_{2}}\} denotes the system of Fourier coefficients of f with respect to \Psi . In the present paper, we investigate the summability properties of Fourier coefficients of functions belonging to the anisotropic Lorentz space. A key tool in our analysis is provided by the grand Lorentz spaces introduced in [ 25 ] . Throughout this paper, the notation A\lesssim B means that A\leq c\,B, for some constant c>0 independent of the essential parameters of the ambient space. The notation A\asymp B signifies a two–sided estimate, that is, both A\lesssim B and B\lesssim A hold. In the definitions below, we use the symbols \bar{p}=(p_{1},p_{2}),\qquad\bar{q}=(q_{1},q_{2})\in\mathbb{R}_{+}^{2}, with all parameters considered in this generality. If 0<p_{1},p_{2}\leq\infty,\qquad\frac{1}{p_{1}}+\frac{1}{p_{1}^{\prime}}=1,\qquad\frac{1}{p_{2}}+\frac{1}{p_{2}^{\prime}}=1, then we write \bar{0}<\bar{p}\leq\bar{\infty} and \frac{1}{\bar{p}}+\frac{1}{\bar{p}^{\prime}}=\bar{1}. Moreover, if p_{1}\leq q_{1} and p_{2}\leq q_{2} , we write \bar{p}\leq\bar{q} . 2. Anisotropic Grand Lorentz Function Spaces Definition 2.1 . [ 26 ] Let \bar{\theta}=(\theta_{1},\theta_{2}) , -\infty<\bar{\theta}<\infty , \bar{p}=(p_{1},p_{2}) , 0<\bar{p}\leq\infty , \bar{q}=(q_{1},q_{2}) , 0<\bar{q}\leq\infty . The anisotropic grand Lorentz space G^{\bar{\theta}}L_{\bar{p},\bar{q}}[0,1]^{2} is defined as the set of all measurable functions f for which the following quasi-norms are finite. For 0<\bar{q}<\infty : \|f\|_{G^{\bar{\theta}}L_{\bar{p},\bar{q}}[0,1]^{2}}=\begin{cases}\displaystyle\sup_{0<\varepsilon_{1},\varepsilon_{2}\leq 1}\varepsilon_{1}^{\theta_{1}}\varepsilon_{2}^{\theta_{2}}\left(\int_{0}^{1}\left(\int_{0}^{1}\left(t_{1}^{\frac{1}{p_{1}}+\varepsilon_{1}}t_{2}^{\frac{1}{p_{2}}+\varepsilon_{2}}f^{*1,*2}(t_{1},t_{2})\right)^{q_{1}}\frac{dt_{1}}{t_{1}}\right)^{\frac{q_{2}}{q_{1}}}\frac{dt_{2}}{t_{2}}\right)^{\frac{1}{q_{2}}},\\ \text{for }\theta_{1}\geq 0,\ \theta_{2}\geq 0,0<p_{i}\leq\infty,i=1,2,\\[8.61108pt] \displaystyle\inf_{\begin{subarray}{c}0<\varepsilon_{1}\leq\frac{1}{p_{1}}\\ 0<\varepsilon_{2}\leq\frac{1}{p_{2}}\end{subarray}}\varepsilon_{1}^{\theta_{1}}\varepsilon_{2}^{\theta_{2}}\left(\int_{0}^{1}\left(\int_{0}^{1}\left(t_{1}^{\frac{1}{p_{1}}-\varepsilon_{1}}t_{2}^{\frac{1}{p_{2}}-\varepsilon_{2}}f^{*1,*2}(t_{1},t_{2})\right)^{q_{1}}\frac{dt_{1}}{t_{1}}\right)^{\frac{q_{2}}{q_{1}}}\frac{dt_{2}}{t_{2}}\right)^{\frac{1}{q_{2}}},\\ \text{for }\theta_{1}<0,\ \theta_{2}<0,0<p_{i}\leq\infty,i=1,2.\end{cases} Remark 2.2 . Unlike the classical anisotropic Lorentz spaces L_{\bar{p},\bar{q}}[0,1]^{2} the grand anisotropic Lorentz spaces G^{\bar{\theta}}L_{\bar{p},\bar{q}}[0,1]^{2} have meaning in cases p_{i}=\infty,\theta_{i}>0 . Remark 2.3 . These spaces were introduced and their properties studied in the works [ 25 ] – [ 26 ] . In case when \bar{\theta}=(0,0) , we have {G^{\bar{\theta}}L_{\bar{p},\bar{q}}[0,1]^{2}}=L_{\bar{p},\bar{q}}[0,1]^{2} . Hence, the anisotropic grand Lorentz space includes in it scales the anisotropic Lorentz space, and the parameter \bar{\theta}=(\theta_{1},\theta_{2}) makes the associated scale of grand spaces more flexible. The grand anisotropic Lorentz spaces allow us to consider the scale of spaces {G^{\bar{\theta}}L_{\bar{p},\bar{q}}[0,1]^{2}} . For \bar{\theta}>\overline{0} the following property holds: (2.1) {G^{-\bar{\theta}}L_{\bar{p},\bar{q}}[0,1]^{2}}.\hookrightarrow L_{\bar{p},\bar{q}}[0,1]^{2}\hookrightarrow{G^{\bar{\theta}}L_{\bar{p},\bar{q}}[0,1]^{2}}. Lemma 2.4 . [ 26 ] Let 0<p_{1},p_{2}<\infty . Then for \theta_{1},\theta_{2}>0 \|f\|_{G^{\bar{\theta}}L_{\bar{p},\bar{\infty}}[0,1]^{2}}\asymp\sup_{t_{1},t_{2}>0}\frac{t_{1}^{\frac{1}{p_{1}}}\cdot t_{2}^{\frac{1}{p_{2}}}}{\left|\ln t_{1}\right|^{\theta_{1}}\left|\ln t_{2}\right|^{\theta_{2}}}f^{*_{1},*_{2}}\left(t_{1},t_{2}\right) \displaystyle\|f\|_{G^{\bar{\theta}}L_{\bar{p},\bar{q}}[0,1]^{2}} \displaystyle\lesssim\left(\int_{0}^{1}\left(\int_{0}^{1}\left(\frac{t_{1}^{\frac{1}{p_{1}}}\cdot t_{2}^{\frac{1}{p_{2}}}}{\left|\ln t_{1}\right|^{\theta_{1}}\left|\ln t_{2}\right|^{\theta_{2}}}f^{*_{1},*_{2}}\left(t_{1},t_{2}\right)\right)^{q_{1}}\frac{dt_{1}}{t_{1}}\right)^{\frac{q_{2}}{q_{1}}}\frac{dt_{2}}{t_{2}}\right)^{q_{2}} \displaystyle\|f\|_{G^{-\bar{\theta}}L_{\bar{p},\bar{q}}[0,1]^{2}} \displaystyle\gtrsim\left(\int_{0}^{1}\left(\int_{0}^{1}\left(t_{1}^{\frac{1}{p_{1}}}\left|\ln t_{1}\right|^{\theta_{1}}t_{2}^{\frac{1}{p_{2}}}\left|\ln t_{2}\right|^{\theta_{2}}f^{*_{1},*_{2}}\left(t_{1},t_{2}\right)\right)^{q_{1}}\frac{dt_{1}}{t_{1}}\right)^{\frac{q_{2}}{q_{1}}}\frac{dt_{2}}{t_{2}}\right)^{q_{2}} From the definition of G^{\bar{\theta}}L_{\bar{p},\bar{q}}[0,1]^{2} and well-known results on Lorentz spaces it is easy to prove the following properties [ 26 ] : (P.1) if \bar{\theta}\leqslant\bar{s} , then G^{\bar{\theta}}L_{\bar{p},\bar{q}}[0,1]^{2}\hookrightarrow G^{\bar{s}}L_{\bar{p},\bar{q}}[0,1]^{2} , (P.2) if \bar{q}<\bar{r} , then G^{\bar{\theta}}L_{\bar{p},\bar{q}}[0,1]^{2}\hookrightarrow G^{\bar{\theta}}L_{\bar{p},\bar{r}}[0,1]^{2} , (P.3) if \bar{p}<\bar{p}_{1},0<\bar{q},\bar{q}_{1}\leq\infty , then G^{\bar{\theta}}L_{\bar{p_{1}},\bar{q_{1}}}[0,1]^{2}\hookrightarrow G^{\bar{\theta}}L_{\bar{p},\bar{q}}[0,1]^{2} , (P.4) if \overline{0}<\bar{\delta}<\overline{1} and \bar{\theta}>\overline{0} , then \|f\|_{G^{\bar{\theta}}L_{\bar{p},\bar{q}}[0,1]^{2}}\asymp\sup_{\begin{subarray}{c}0<\varepsilon_{1}\leqslant\delta_{1}\\ 0<\varepsilon_{2}\leqslant\delta_{2}\end{subarray}}\varepsilon_{1}^{\theta_{1}}\varepsilon_{2}^{\theta_{2}}\left(\int_{0}^{1}\left(\int_{0}^{1}\left(t_{1}^{\frac{1}{p_{1}}+\varepsilon_{1}}t_{2}^{\frac{1}{p_{2}}+\varepsilon_{2}}f^{*_{1},*_{2}}\left(t_{1},t_{2}\right)\right)^{q_{1}}\frac{dt_{1}}{t_{1}}\right)^{\frac{q_{2}}{q_{1}}}\frac{dt_{2}}{t_{2}}\right)^{q_{2}} and, if 0<\bar{\delta}<1/\bar{p} , \|f\|_{G^{-\bar{\theta}}L_{\bar{p},\bar{q}}[0,1]^{2}}\asymp\inf_{\begin{subarray}{c}0<\varepsilon_{1}<\delta_{1}\\ 0<\varepsilon_{2}<\delta_{2}\end{subarray}}\varepsilon_{1}^{-\theta_{1}}\varepsilon_{2}^{-\theta_{2}}\left(\int_{0}^{1}\left(\int_{0}^{1}\left(t_{1}^{\frac{1}{p_{1}}-\varepsilon_{1}}t_{2}^{\frac{1}{p_{2}}-\varepsilon_{2}}f^{*_{1},*_{2}}\left(t_{1},t_{2}\right)\right)^{q_{1}}\frac{dt_{1}}{t_{1}}\right)^{\frac{q_{2}}{q_{1}}}\frac{dt_{2}}{t_{2}}\right)^{q_{2}} (P.5) Let 0<\bar{p}\leq\infty,\bar{\theta}<\bar{\lambda} and \bar{\theta}-\frac{1}{\bar{q}} then the following embedding holds: G^{\bar{\theta}}L_{{\bar{p}},\bar{\tau}}[0,1]^{2}\hookrightarrow G^{\bar{\lambda}}L_{{\bar{p}},\bar{\tau}}[0,1]^{2} , (P.6)