The Properties of Cesáro General Fourier Sums of Functions with Derivatives of Lipschitz Class Functions

G. Tutberidze, V. Tsagareishvili, and G. Cagareishvil · 2025

Read The Properties of Cesáro General Fourier Sums of Functions with Derivatives of Lipschitz Class Functions by G. Tutberidze, V. Tsagareishvili, and G. Cagareishvil (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-2509-07466.txt.

Introduction

G. Tutberidze and 2 co-authors study Cesáro means of Fourier series in a 2025 paper. A Fourier series rebuilds a function from waves in an orthonormal system, a set of building blocks at right angles to each other. Cesáro means average the partial sums. That averaging often tames a series that will not settle on its own. Yet smoothness of a function alone does not ensure its series can be summed for a general system.

The authors ask what extra conditions on the system functions make the Cesáro means stay bounded for all functions of one class. In that class, the derivative of the function lies in the Lipschitz class Lip 1, so it cannot change too fast. They find such conditions and show they are sharp. Sharp means the conditions cannot be weakened. Results like these tell analysts when averaging can rescue a Fourier expansion.

Catalog record

CREATORG. Tutberidze, V. Tsagareishvili, and G. Cagareishvil
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE26 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2509-07466

Browse the full library · About the Tower of Babel

Full text

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

The Properties of Cesáro General Fourier Sums of Functions with Derivatives of Lipschitz Class Functions Thanks: The research of third author is supported by Shota Rustaveli National Science Foundation grant no. FR-24-698. G. Tutberidze, V. Tsagareishvili, and G. Cagareishvil Abstract . In this paper, we investigate the Cesáro means of Fourier series with respect to general orthonormal systems (ONS), when the function f f belongs to a certain differentiable class of functions. It is well known that the membership of a function f ≢ 0 f\not\equiv 0 in a differentiable class does not, in general, guarantee the summability of its Fourier series with respect to an arbitrary ONS. Therefore, in order for the Fourier series with respect to a given ONS to be summable, one must impose additional conditions on the system functions { φ n } \{\varphi_{n}\} . The main objective of this work is to determine such conditions on the functions φ n \varphi_{n} of the ONS under which the Cesáro means of the Fourier series of any function whose derivative belongs to the Lipschitz class Lip 1 \mathrm{Lip}_{1} are uniformly bounded. The results obtained are sharp in the sense that the conditions cannot be essentially weakened. 2010 Mathematics Subject Classification. 42C10, 46B07 Key words and phrases: General Fourier series, Fourier coefficients, Cesáro sums, Lipschitz class, Differentiable functions, ONS, Banach space. 1. Preliminary Definitions and Theorems Let C_{L} be the class of functions such that \frac{df}{dx}\in\mathrm{Lip}_{1} . The space \mathrm{Lip}_{1} is a Banach space with the norm \|f\|_{\mathrm{Lip}_{1}}=\|f\|_{C}+\max_{x\neq y}\frac{|f(x)-f(y)|}{|x-y|}, where C[0,1] is a space of continuous functions, with the following norm: \left\|f\right\|_{c}=\max_{x\in[0,1]}\left|f\left(x\right)\right|. Suppose that (\varphi_{n}) is an ONS on [0,1] , and let f\in L_{2} be an arbitrary function. The numbers C_{n}(f)=\int_{0}^{1}f(x)\varphi_{n}(x)\,dx,\quad n=1,2,\dots are called the Fourier coefficients of the function f with respect to the ONS (\varphi_{n}) . The Fourier series of the function f is given by the expression: \sum_{k=1}^{\infty}C_{k}(f)\,\varphi_{k}(x). (1.1) The general partial sum of the series ( 1.1 ), denoted by S_{n}(x,f) , is given by the expression S_{n}(x,f)=\sum_{k=1}^{n}C_{k}(f)\,\varphi_{k}(x). (1.2) The Cesáro means of order \alpha of the series ( 1.1 ), denoted by \sigma_{n}^{\alpha}(x,f) (see [ 1 ] , Ch. 2, §1, p. 73; §6, p. 115), are given by the expression \sigma_{n}^{\alpha}(x,f)=\frac{1}{A_{n}^{\alpha}}\sum_{k=0}^{n-1}A_{n-k}^{\alpha}\,a_{k}\,\varphi_{k}(x), where \alpha>0, and A_{n}^{\alpha}=\frac{(1+\alpha)(2+\alpha)\cdots(n+\alpha)}{n!}. For \alpha>0 , the following estimate holds: A_{1}n^{\alpha}<A_{n}^{\alpha}<A_{2}n^{\alpha}. We define the function g_{n}(x) by g_{n}(x)=\int_{0}^{x}\varphi_{n}(u)\,du, and define Q_{n}(u,x) as Q_{n}(u,x)=\frac{1}{A_{n}^{\alpha}}\sum_{k=0}^{n-1}A_{n-k}^{\alpha}\,g_{k}(u)\,\varphi_{k}(x). Let us denote D_{nk}^{\alpha}=\frac{A_{n}^{\alpha}}{A_{n-k}^{\alpha}}. Then, the function Q_{n}(u,x) can be represented as Q_{n}(u,x)=\sum_{k=0}^{n-1}D_{nk}^{\alpha}g_{k}(u)\varphi_{k}(x). (1.3) Finally, we define H_{n}(x)=\sum_{i=1}^{n-1}\left|\int_{0}^{\frac{i}{n}}Q_{n}(u,x)\,du\right|, (1.4) Lemma 1.1 . If (\varphi_{k}) is an arbitrary ONS, then \frac{1}{n^{2}}\sum_{k=1}^{n}\varphi_{k}^{2}(x)=O(n^{-\frac{1}{2}}),\quad\text{a.e. on }[0,1]. (1.5) Indeed, since \frac{1}{n^{2}}\sum_{k=1}^{n}\varphi_{k}^{2}(x)\leq\frac{1}{\sqrt{n}}\sum_{k=1}^{n}k^{-\frac{3}{2}}\varphi_{k}^{2}(x), and \frac{1}{\sqrt{n}}\sum_{k=1}^{\infty}k^{-\frac{3}{2}}\int_{0}^{1}\varphi_{k}^{2}(x)\,dx=\frac{1}{\sqrt{n}}\sum_{k=1}^{\infty}k^{-\frac{3}{2}}<+\infty, it follows from Levi’s theorem that the convergence holds almost everywhere, which proves ( 1.5 ). Notation: Suppose G=[0,1]\setminus F , where F=\left\{x\in[0,1]\,\middle|\,\lim_{n\to\infty}\frac{1}{n^{2}}\sum_{k=1}^{n}\varphi_{k}^{2}(x)=\infty\right\}. It is easy to show that |F|=0\quad\text{and}\quad|G|=1. Definition 1.1 . We say that f\in E(x) if, at the point x\in[0,1] , \limsup_{n\to\infty}\left|\sigma_{n}^{\alpha}(x,f)\right|<+\infty. Lemma 1.2 . For every integer i with 1\leq i\leq n and for all x\in[0,1] , the following inequality (see ( 1.3 )): \int_{\frac{i-1}{n}}^{\frac{i}{n}}|Q_{n}(u,x)|\,du\leq\left(\frac{1}{n^{2}}\sum_{k=1}^{n}\varphi_{k}^{2}(x)\right)^{\frac{1}{2}}. (1.6) Proof. By the Bessel inequality, it follows that \sum_{k=1}^{\infty}g_{k}^{2}(u)=\sum_{k=1}^{\infty}\left(\int_{0}^{u}\varphi_{k}(t)\,dt\right)^{2}\leq\int_{0}^{u}dt\leq 1. (1.7) Here, the function g_{k} is defined by g_{k}(u)=\int_{0}^{u}\varphi_{k}(t)\,dt,\quad u\in[0,1]. By applying the Cauchy and Hölder inequalities, as well as inequality ( 1.7 ), and taking into account that D_{nk}^{\alpha}<1 , we conclude, for each i=1,\dots,n , \displaystyle\int_{\frac{i-1}{n}}^{\frac{i}{n}}|Q_{n}(u,x)|\,du\leq\frac{1}{\sqrt{n}}\left(\int_{\frac{i-1}{n}}^{\frac{i}{n}}Q_{n}^{2}(u,x)\,du\right)^{\frac{1}{2}} \displaystyle=\frac{1}{\sqrt{n}}\left(\int_{\frac{i-1}{n}}^{\frac{i}{n}}\left(\sum_{k=1}^{n}D_{nk}^{\alpha}g_{k}(u)\varphi_{k}(x)\right)^{2}du\right)^{\frac{1}{2}} \displaystyle\leq\frac{1}{\sqrt{n}}\left(\int_{\frac{i-1}{n}}^{\frac{i}{n}}\sum_{k=1}^{n}(D_{nk}^{\alpha})^{2}g_{k}^{2}(u)\,du\cdot\sum_{k=1}^{n}\varphi_{k}^{2}(x)\right)^{\frac{1}{2}} \displaystyle\leq\frac{1}{n}\left(\sum_{k=1}^{n}\varphi_{k}^{2}(x)\right)^{\frac{1}{2}}. Hence, Lemma 1.2 is proved. ∎ Lemma 1.3 . Let (\varphi_{n}) be an ONS on [0,1] , and let f\in C_{L} . Then \sigma_{n}^{\alpha}(x,f)=f(1)\,\sigma_{n}^{\alpha}(x,q)-\int_{0}^{1}f^{\prime}(u)\,Q_{n}(u,x)\,du, (1.8) where q(u)=1 for all u\in[0,1] . Proof. Integrating by parts, we have C_{n}(f)=\int_{0}^{1}f(u)\,\varphi_{n}(u)\,du=f(1)\int_{0}^{1}\varphi_{n}(u)\,du-\int_{0}^{1}f^{\prime}(u)\,g_{n}(u)\,du, that is, C_{n}(f)=f(1)C_{n}(q)-\int_{0}^{1}f^{\prime}(u)\,g_{n}(u)\,du. (1.9) It follows that, \sum_{k=0}^{n-1}D_{nk}^{\alpha}C_{k}(f)\,\varphi_{k}(x)=f(1)\sum_{k=0}^{n-1}D_{nk}^{\alpha}C_{k}(q)\,\varphi_{k}(x)-\int_{0}^{1}f^{\prime}(u)\sum_{k=0}^{n-1}D_{nk}^{\alpha}g_{k}(u)\varphi_{k}(x)\,du. By taking equation ( 1.3 ) into consideration, we obtain the desired result ( 1.8 ). ∎ Theorem 1.1 (S. Banach [ 2 ] ) . For any f\in L_{2} , \left(f\not\sim 0\right) , there exists an ONS (\varphi_{n}) such that \limsup_{n\to\infty}\left|\sigma_{n}^{\alpha}(x,f)\right|=+\infty,\quad\text{a.e. on }[0,1]. Theorem 1.2 (see [ 3 ] ) . If f,F\in L_{2} , then the following identity holds: \displaystyle\int_{0}^{1}f^{\prime}(u)F(u)\,du \displaystyle=n\sum_{i=1}^{n-1}\left(\int_{\frac{i-1}{n}}^{\frac{i}{n}}\left(f(u)-f\left(u+\frac{1}{n}\right)\right)\,du\int_{0}^{\frac{i}{n}}F(u)\,du\right) \displaystyle\quad+n\sum_{i=1}^{n-1}\int_{\frac{i-1}{n}}^{\frac{i}{n}}\int_{\frac{i-1}{n}}^{\frac{i}{n}}(f(u)-f(v))\,dv\,F(u)\,du \displaystyle\quad+n\int_{1-\frac{1}{n}}^{1}f(u)\,du\int_{0}^{1}F(u)\,du. (1.10) 2. Statement of the Main Problem As is well known, the convergence or summability of Fourier series with respect to classical ONS – such as the trigonometric system (see [ 4 , Ch. 4, #1, p. 150] ), the Haar system (see [ 4 , Ch. 3, #1, p. 70] ), and the Walsh system (see [ 4 , Ch. 4, #5, p. 150] ) – for differentiable functions is a relatively straightforward matter. This is primarily due to the well-understood structure and good localization properties of these systems, which align well with the smoothness of the functions under consideration. However, the situation is markedly different when dealing with general ONS. In such cases, classical techniques may fail to guarantee convergence, and even summability can become a subtle and complex issue. The lack of smoothness, poor localization, or irregularity in the structure of the system often leads to new challenges, requiring refined tools and more delicate analysis to obtain meaningful results. Indeed, as shown by Banach’s classical result [ 2 ] , the Fourier series with respect to certain ONS may fail to be summable even for simple functions such as f(x)=1 on the interval [0,1] . This highlights the fact that, in general, summability cannot be taken for granted, even in the case of very regular or constant functions. In this paper, we identify those ONS (\varphi_{n}) for which the condition \sigma_{n}^{\alpha}(x,f)=O(1) holds for every function f whose derivative f^{\prime} belongs to the Lipschitz class \mathrm{Lip}1 , at a given point x\in G . Similar questions have been investigated in a number of recent works (see [ 6 , 7 , 10 , 11 , 12 ] ), where related aspects of summability and convergence for various ONS were considered. 3. Main Results Theorem 3.1 . Let (\varphi_{n}) be an ONS on the interval [0,1] , and let p,q\in E(x) , where p(u)=u and q(u)=1 for all u\in[0,1] . If, for a point x\in G , the condition H_{n}(x)=O(1) (3.1) holds, then for any function f\in C_{L} , the following estimate is valid at the point x\in G : \sigma_{n}^{\alpha}(x,f)=O(1). Proof. If we substitute F(x)=Q_{n}(u,x) and f=f^{\prime} into equation ( 1.10 ), we obtain \displaystyle\int_{0}^{1}f^{\prime}(u)Q_{n}(u,x)\,du \displaystyle=n\sum_{i=1}^{n-1}\left(\int_{\frac{i-1}{n}}^{\frac{i}{n}}\left(f^{\prime}(u)-f^{\prime}\left(u+\frac{1}{n}\right)\right)du\int_{0}^{\frac{i}{n}}Q_{n}(u,x)\,du\right) \displaystyle\quad+n\sum_{i=1}^{n-1}\int_{\frac{i-1}{n}}^{\frac{i}{n}}\int_{\frac{i-1}{n}}^{\frac{i}{n}}\left(f^{\prime}(u)-f^{\prime}(v)\right)dv\,Q_{n}(u,x)\,du \displaystyle\quad+n\int_{1-\frac{1}{n}}^{1}f^{\prime}(u)\,du\int_{0}^{1}Q_{n}(u,x)\,du. (3.2) Due to ( 3.1 ) and the fact that f^{\prime}\in Lip_{1}, we obtain \displaystyle|I_{1}| \displaystyle=n\cdot O\left(\tfrac{1}{n}\right)\sum_{i=1}^{n-1}\int_{\frac{i-1}{n}}^{\frac{i}{n}}du\left|\int_{0}^{\frac{i}{n}}Q_{n}(u,x)\,du\right| \displaystyle=\frac{O(1)}{n}\sum_{i=1}^{n-1}\left|\int_{0}^{\frac{i}{n}}Q_{n}(u,x)\,du\right|=O(1)T_{n}(x)=O(1). Next, since f^{\prime}\in\mathrm{Lip}_{1} and x\in G , by applying inequality ( 1.7 ) from Lemma 1.1 , which proves that D_{nk}^{\alpha}<1 , we obtain: \displaystyle|I_{2}| \displaystyle\leq n\cdot O\left(\tfrac{1}{n^{2}}\right)\sum_{i=1}^{n}\int_{\frac{i-1}{n}}^{\frac{i}{n}}|Q_{n}(u,x)|\,du \displaystyle=O(1)\cdot\tfrac{1}{n}\int_{0}^{1}|Q_{n}(u,x)|\,du=O(1)\cdot\tfrac{1}{n}\left(\int_{0}^{1}Q_{n}^{2}(u,x)\,du\right)^{\frac{1}{2}} \displaystyle=O(1)\cdot\tfrac{1}{n}\left(\int_{0}^{1}\left(\sum_{k=1}^{n}D_{nk}^{\alpha}g_{k}(u)\varphi_{k}(x)\right)^{2}du\right)^{\frac{1}{2}} \displaystyle=O(1)\cdot\tfrac{1}{n}\left(\sum_{k=1}^{n}(D_{nk}^{\alpha})^{2}\varphi_{k}^{2}(x)\int_{0}^{1}g_{k}^{2}(u)\,du\right)^{\frac{1}{2}} \displaystyle\leq O(1)\cdot\tfrac{1}{n}\left(\sum_{k=1}^{n}\varphi_{k}^{2}(x)\right)^{\frac{1}{2}}=O(1)\cdot\left(\tfrac{1}{n^{2}}\sum_{k=1}^{n}\varphi_{k}^{2}(x)\right)^{\frac{1}{2}}=O(1). Furthermore, according to equation ( 1.8 ), if F(u)=p(u)=u,\quad u\in[0,1], then we get \sigma_{n}^{\alpha}(x,p)=f(1)\,\sigma_{n}^{\alpha}(x,q)-\int_{0}^{1}Q_{n}(u,x)\,du Now, according to the conditions of Theorem 1.1 , where p,q\in E(x) or \sigma_{n}^{\alpha}(x,p)=O(1)\text{ and }\sigma_{n}^{\alpha}(x,q)=O(1), we derive \left|\int_{0}^{1}Q_{n}(u,x)\,du\right|=O(1). Noting that f^{\prime}(u)\in\mathrm{Lip}_{1} , the term I_{3} can be estimated as |I_{3}|\leq n\left|\int_{1-\frac{1}{n}}^{1}f^{\prime}(u)\,du\right|\cdot\left|\int_{0}^{1}R_{n}(u,x)\,du\right|=O(1). Therefore, given that |I_{1}|=O(1) , |I_{2}|=O(1) , and |I_{3}|=O(1) , it follows from equation ( 3.2 ) that \left|\int_{0}^{1}f^{\prime}(u)Q_{n}(u,x)\,du\right|=O(1) is fulfilled. Once again, using equation ( 1.8 ) for f\in C_{L} , it follows that \sigma_{n}^{\alpha}(x,f)=f(1)\,\sigma_{n}^{\alpha}(x,q)-\int_{0}^{1}f^{\prime}(u)Q_{n}(u,x)\,du, where p(u)=u and q(u)=1 , for u\in[0,1] . From the last equality, we have \sigma_{n}^{\alpha}(x,f)=O(1). Hence, Theorem 3.1 is proved. ∎ Theorem 3.2 . Let (\varphi_{n}) be an ONS on the interval [0,1]

Read the complete text