Generalized Bohr inequalities for K-quasiconformal harmonic mappings and their applications

Raju Biswas and Rajib Mandal

arXiv: 2411.01837v1 · PDF · 4 November 2024
Categories: math.CV
Journal: Ukr. Math. J. 77 (2025), 1611-1634
DOI: 10.1007/s11253-025-02543-8
Licence: CC BY 4.0 The site shows the text.

Abstract

The classical Bohr theorem and its subsequent generalizations have become active areas of research, with investigations conducted in numerous function spaces. Let \{\psi_n(r)\}_{n=0}^\infty be a sequence of non-negative continuous functions defined on [0,1) such that the series \sum_{n=0}^\infty \psi_n(r) converges locally uniformly on the interval [0, 1). The main objective of this paper is to establish several sharp versions of generalized Bohr inequalities for the class of K-quasiconformal sense-preserving harmonic mappings on the unit disk \D := \{z \in \mathbb{C} : |z| < 1\}. To achieve these, we employ the sequence of functions \{\psi_n(r)\}_{n=0}^\infty in the majorant series rather than the conventional dependence on the basis sequence \{r^n\}_{n=0}^\infty. As applications, we derive a number of previously published results as well as a number of sharply improved and refined Bohr inequalities for harmonic mappings in \D. Moreover, we obtain a convolution counterpart of the Bohr theorem for harmonic mapping within the context of the Gaussian hypergeometric function

Items 40

Theorem A

Theorem A A

ZJ0K

[1KP2017] Let f(z)=\sum_{n=0}^{\infty} a_nz^n be analytic in \mathbb{D} and |f(z)|\leq 1. Then

\begin{aligned} |f(z)|+\sum_{n=N}^{\infty}|a_n||z|^n\leq 1\end{aligned}

for |z|= r\leq R_N, where R_N\in(0,1) is the positive root of the equation \psi_N( r)=0, \psi_N( r)=2(1+ r) r^N-(1- r)^2. The radius R_N is the best possible. Moreover,

\begin{aligned}|f(z)|^2+\sum_{n=N}^{\infty}|a_n||z|^n\leq 1\end{aligned}

for |z|= r\leq R_N', where R_N'\in (0,1) is the positive root of the equation (1+ r) r^N-(1- r)^2=0. The radius R_N' is the best possible.

Theorem B

Theorem B B

T52O

[LSX2018] Let f(z)=\sum_{n=0}^{\infty} a_nz^n be analytic in \mathbb{D} and |f(z)|\leq 1. Then

\begin{aligned} |f(z)|+|f'(z)||z|+\sum_{n=2}^{\infty}|a_n||z|^n\leq 1\quad\text{for}\quad |z|=r\leq (\sqrt{17}-3)/4.\end{aligned}

The radius (\sqrt{17}-3)/4 is the best possible.

Theorem C

Theorem C C

PJEL

[1KP2018] Let f(z)=\sum_{n=0}^{\infty} a_nz^n be analytic in \mathbb{D}, |f(z)|\leq 1 and S_r(f) denotes the area of the image of the subdisk |z|<r under mapping f. Then

\begin{aligned} \sum_{n=0}^{\infty} |a_n|r^n+\frac{16}{9}\left(\frac{S_r(f)}{\pi}\right)\leq 1\quad\text{for}\quad r\leq\frac{1}{3}.\end{aligned}

The numbers 1/3, 16/9 cannot be improved. Moreover,

\begin{aligned} |a_0|^2+\sum_{n=1}^{\infty} |a_n| r^n+\frac{9}{8}\left(\frac{S_r(f)}{\pi}\right)\leq 1\quad\text{for}\quad r\leq\frac{1}{2},\end{aligned}

and the numbers 1/2, 9/8 cannot be improved.

Theorem D

Theorem D D

3894

[PVW2020] Let f(z)=\sum_{n=0}^{\infty} a_nz^n be analytic in \mathbb{D} and |f(z)|\leq 1. Then,

\begin{aligned} \sum_{n=0}^\infty |a_n| r^n+\left(\frac{1}{1+|a_0|}+\frac{ r}{1- r}\right)\sum_{n=1}^\infty |a_n|^2 r^{2n}\leq 1\quad\text{for}\quad r\leq 1/(2+|a_0|),\end{aligned}

and the numbers 1/(1+|a_0|) and 1/(2+|a_0|) cannot be improved. Moreover,

\begin{aligned} |a_0|^2+\sum_{n=1}^\infty |a_n| r^n+\left(\frac{1}{1+|a_0|}+\frac{ r}{1- r}\right)\sum_{n=1}^\infty |a_n|^2 r^{2n}\leq 1\quad\text{for}\quad r\leq 1/2,\end{aligned}

and the numbers 1/(1+|a_0|) and 1/2 cannot be improved.

Theorem E

Theorem E E

OK1P

[LLP2021] Let f(z)=\sum_{n=0}^{\infty} a_nz^n be analytic in \mathbb{D} and |f(z)|\leq 1. Then,

\begin{aligned} |f(z)|+|f'(z)|r+\sum_{n=2}^\infty |a_n| r^n+\left(\frac{1}{1+|a_0|}+\frac{r}{1- r}\right)\sum_{n=1}^\infty |a_n|^2 r^{2n}\leq 1\end{aligned}

for r\leq (\sqrt{17}-3)/4. The number (\sqrt{17}-3)/4 is the best possible. Moreover,

\begin{aligned} |f(z)|^2+|f'(z)|r+\sum_{n=2}^\infty |a_n| r^n+\left(\frac{1}{1+|a_0|}+\frac{r}{1- r}\right)\sum_{n=1}^\infty |a_n|^2 r^{2n}\leq 1\end{aligned}

for r\leq r_0, where r_0\approx 0.385795 is the unique positive root of the equation 1-2r-r^2-r^3-r^4=0 and the number r_0 is the best possible.

Theorem F

Theorem F F

PZQ9

[KKP2022] Suppose that f(z)=\sum_{n=0}^\infty a_n z^n is an analytic function in \mathbb{D} with |f(z)|\leq 1 and p\in(0,2]. If \{\psi_n(r)\}_{n=1}^\infty\in{\mathcal{F}} satisfies the inequality

\begin{aligned} \frac{2}{p}\sum_{n=1}^\infty \psi_n(r)<\psi_0(r)\quad\text{for}\quad r\in[0,R),\end{aligned}

where R is minimal positive root of the equation (2/p)\sum_{n=1}^\infty \psi_n(r)=\psi_0(r), then the following inequality holds

\begin{aligned} |a_0|^p\psi_0(r)+\sum_{n=1}^\infty |a_n| \psi_n(r)\leq \psi_0(r)\quad\text{for}\quad r\leq R.\end{aligned}

In the case when (2/p)\sum_{n=1}^\infty \psi_n(r)>\psi_0(r) in some interval (R, R+\varepsilon), then the number R cannot be improved. If the functions \psi_k(r) (k\geq 0) are smooth functions, then the last condition is equivalent to the inequality (2/p)\sum_{n=1}^\infty \psi_n'(R)>\psi_0'(R).

Theorem G

Theorem G G

HXWJ

[PVW2021] Suppose that f(z)=\sum_{n=0}^\infty a_n z^n is an analytic function in \mathbb{D} with |f(z)|\leq 1 and p\in(0,2]. If \{\psi_n(r)\}_{n=1}^\infty\in{\mathcal{F}} satisfies the inequality

\begin{aligned} \frac{2}{p}\sum_{n=1}^\infty \psi_n(r)<\psi_0(r)\quad\text{for}\quad r\in[0,R),\end{aligned}

where R is minimal positive root of the equation 2\sum_{n=1}^\infty \psi_n(r)=p\psi_0(r), then the following inequality holds

\begin{aligned} |a_0|^p\psi_0(r)+\sum_{n=1}^\infty |a_n| \psi_n(r)+\sum_{n=1}^\infty |a_n|^2\left(\frac{\psi_{2n}(r)}{1+|a_0|}+\Psi_{2n+1}(r)\right)\leq \psi_0(r)\quad\text{for}\quad r\leq R,\end{aligned}

where \Psi_t(r)=\sum_{k=t}^\infty \psi_k(r). In the case when 2\sum_{n=1}^\infty \psi_n(r)>p\psi_0(r) in some interval (R, R+\varepsilon), then the number R cannot be improved.

Question

Question 1.1

IYB8

Instead of the conventional dependence on the basis sequence \{r^n\}_{n=0}^\infty, is it possible to establish general results by using the sequence \{\psi_n(r)\}_{n=0}^\infty\in{\mathcal{F}} in the majorant series for harmonic mappings in the settings of Theorems B-E?

Question

Question 1.2

8U7J

Can we establish the harmonic extension of the generalized Bohr inequality in the context of Theorem F?

Question

Question 1.3

T6GG

Can we establish the harmonic extension of the generalized Bohr inequality in refined form in the context of Theorem G?

Question

Question 1.4

KZPK

Can we establish several sharp versions of generalized Bohr inequalities for harmonic mappings in the context of Theorems F and G?

Lemma

Lemma 2.1

EV29

[K2006, Pick’s invariant form of Schwarz’s lemma] Suppose f is analytic in \mathbb{D} with |f(z)|\leq1, then

\begin{aligned} |f(z)|\leq \frac{|f(0)|+|z|}{1+|f(0)||z|}\quad\text{for}\quad z\in\mathbb{D}.\end{aligned}

Used by Proof, Proof

Lemma

Lemma 2.2

O0S2

[DP2008] Suppose f is analytic in \mathbb{D} with |f(z)|\leq1, then we have

\begin{aligned} \frac{\left|f^{(n)}(z)\right|}{n!}\leq \frac{1-|f(z)|^2}{(1-|z|)^{n-1}(1-|z|^2)}\quad\text{and}\quad |a_n|\leq 1-|a_0|^2\quad\text{for}\quad n\geq 1,\;|z|<1.\end{aligned}

Used by Proof, Proof, Proof, Proof

Lemma

Lemma 2.3

PKRM

[C1940][PVW2020, Lemma B] Suppose that f\in\mathcal{B}(\mathbb{D},\mathbb{C}) and f(z) = \sum_{k=0}^\infty a_kz^k. Then the following inequalities hold.

(i) \vert a_{2k+1}\vert\leq 1-\vert a_0\vert^2-\cdots-\vert a_k\vert^2, k=0,1,\ldots
(ii) \vert a_{2k}\vert\leq 1-\vert a_0\vert^2-\cdots-\vert a_{k-1}\vert^2-\frac{\vert a_k\vert^2}{1+\vert a_0\vert}, k=1,2,\ldots.

Further, to have equality in (i) it is necessary that f is a rational function of the form

\begin{aligned} f(z)=\frac{a_0+a_1z+\cdots+a_nz^n+\varepsilon z^{2n+1}}{1+(\overline{a_n}z^n+\cdots+\overline{a_0}z^{2n+1})\varepsilon},\quad|\varepsilon|=1,\end{aligned}

and to have the equality in (ii) it is necessary that f is a rational function of the form

\begin{aligned} f(z)=\frac{a_0+a_1z+\cdots+(a_n/(1+|a_0|))z^n+\varepsilon z^{2n+1}}{1+((\overline{a_n}/(1+|a_0|))z^n+\cdots+\overline{a_0}z^{2n})\varepsilon},\quad|\varepsilon|=1,\end{aligned}

where the term a_0\overline{a_n}^2\varepsilon is non-negative real.

Used by Proof

Lemma

Lemma 2.4

PZ1M

[PVW2021] Let \{\psi_n(r)\}_{n=1}^\infty be a decreasing sequence of non-negative functions in [0,r_\psi), and h, g be analytic in \Bbb{D} such that |g'(z)|\leq k|h'(z)| in \Bbb{D} and for some k\in[0,1], where h(z)=\sum_{n=0}^\infty a_n z^n and g(z)=\sum_{n=0}^\infty b_n z^n. Then,

\begin{aligned} \sum_{n=1}^\infty |b_n|^2 \psi_n(r)\leq k^2\sum_{n=1}^\infty |a_n|^2 \psi_n(r)\quad\text{for}\quad r\in[0,r_\psi).\end{aligned}

Used by Proof, Proof, Proof, Proof

Lemma

Lemma 2.5

4X8V

[PVW2021, Proof of Theorem 1] Suppose f(z)=\sum_{n=0}^\infty a_nz^n is analytic in \mathbb{D} with |f(z)|\leq1 and \{\psi_n(r)\}_{n=1}^\infty \in{\mathcal{F}}. Then, we have the following inequality:

\begin{aligned} \sum_{n=1}^\infty |a_n|\psi_n(r)+\sum_{n=1}^\infty |a_n|^2\left(\frac{\psi_{2n}(r)}{1+|a_0|}+\Psi_{2n+1}(r)\right)\leq \left(1-|a_0|^2\right)\Psi_1(r),\end{aligned}

where \Psi_t(r)=\sum_{k=t}^\infty \psi_k(r).

Lemma

Lemma 2.6

8P5Z

Suppose f(z)=\sum_{n=0}^\infty a_nz^n is analytic in \mathbb{D} with |f(z)|\leq1 and \{\psi_n(r)\}_{n=1}^\infty \in{\mathcal{F}}. Then, we have the following inequality:

\begin{aligned} \sum_{n=2}^\infty |a_n|\psi_n(r)+\sum_{n=1}^\infty |a_n|^2\left(\frac{\psi_{2n}(r)}{1+|a_0|}+\Psi_{2n+1}(r)\right)\leq \left(1-|a_0|^2\right)\Psi_2(r),\end{aligned}

where \Psi_t(r)=\sum_{k=t}^\infty \psi_k(r).

Used by Proof, Proof, Proof

Proof of Lemma 2.6

In view of Lemma 2.3, we have

\begin{aligned}\sum_{n=2}^\infty |a_n|\psi_n(r)&=\sum_{n=1}^\infty |a_{2n}|\psi_{2n}(r)+\sum_{n=1}^\infty |a_{2n+1}|\psi_{2n+1}(r)\\[2mm] &\leq &\sum_{n=1}^\infty \left(1-|a_0|^2-\sum_{k=1}^{n-1}|a_k|^2-\frac{|a_n|^2}{1+|a_0|}\right)\psi_{2n}(r)\\[2mm] &&+\sum_{n=1}^\infty \left(1-|a_0|^2-\sum_{k=1}^{n}|a_k|^2\right)\psi_{2n+1}(r)\\[2mm] &=\left(1-|a_0|^2\right)\sum_{n=2}^\infty \psi_n(r)-\sum_{n=1}^\infty\frac{|a_n|^2}{1+|a_0|}\psi_{2n}(r)\\[2mm] &&-\left(\sum_{n=1}^\infty\left(\sum_{k=1}^{n-1}|a_k|^2\right) \psi_{2n}(r)+\sum_{n=1}^\infty\left(\sum_{k=1}^{n}|a_k|^2\right) \psi_{2n+1}(r)\right).\end{aligned}

It is evident that

\begin{aligned} &&\sum_{n=1}^\infty\left(\sum_{k=1}^{n-1}|a_k|^2\right) \psi_{2n}(r)+\sum_{n=1}^\infty\left(\sum_{k=1}^{n}|a_k|^2\right) \psi_{2n+1}(r)\\[2mm] &=|a_1|^2\sum_{k=3}^\infty \psi_k(r)+|a_2|^2\sum_{k=5}^\infty \psi_k(r)+|a_3|^2\sum_{k=7}^\infty \psi_k(r)+\cdots\\[2mm] &=\sum_{n=1}^\infty |a_n|^2\left(\sum_{k=2n+1}^\infty \psi_k(r)\right).\end{aligned}

Therefore, we have

\begin{aligned} \sum_{n=2}^\infty |a_n|\psi_n(r)&\leq&\left(1-|a_0|^2\right)\sum_{n=2}^\infty \psi_n(r)-\sum_{n=1}^\infty\frac{|a_n|^2}{1+|a_0|}\psi_{2n}(r)-\sum_{n=1}^\infty |a_n|^2\Psi_{2n+1}\\[2mm] &=\left(1-|a_0|^2\right)\sum_{n=2}^\infty \psi_n(r)-\sum_{n=1}^\infty|a_n|^2\left(\frac{\psi_{2n}(r)}{1+|a_0|}+\Psi_{2n+1}(r)\right).\end{aligned}

This completes the proof.

Uses Lemma 2.3

Theorem

Theorem 3.1

UMEU

Suppose that f(z)=h(z)+\overline{g(z)}=\sum_{n=0}^\infty a_n z^n+\overline{\sum_{n=1}^\infty b_n z^n} is a sense-preserving K-quasiconformal harmonic mapping in \mathbb{D}, where h(z) is bounded in \mathbb{D}. If p\in(0,2] and \{\psi_n(r)\}_{n=1}^\infty\in{\mathcal{F}} is a decreasing sequence and satisfies the inequality

\begin{aligned} \Phi_1(r):=\frac{2K}{K+1}\sum_{n=1}^\infty \psi_n(r)+\sum_{t=1}^N c_t\left(\sum_{n=1}^\infty n \left(\psi_n(r)\right)^2\right)^t-\frac{p}{2}\psi_0(r)\leq 0\end{aligned}

for r\leq R_{N,K}, then the following inequality holds

\begin{aligned} |a_0|^p\psi_0(r)+\sum_{n=1}^\infty |a_n| \psi_n(r)+\sum_{n=1}^\infty |b_n| \psi_n(r)+G\left(\sum_{n=1}^\infty n|a_n|^2(\psi_n(r))^2\right)\leq \psi_0(r)\Vert h(z)\Vert_\infty\end{aligned}

for r\leq R_{N,K}, where G(x) is a polynomial defined in (?) and R_{N,K}\in(0,1) is minimal positive root of the equation \Phi_1(r)=0. In the case when \Phi_1(r)>0 in some interval (R_{N,K}, R_{N,K}+\varepsilon) (\varepsilon>0), then the number R_{N,K} cannot be improved.

Used by Remark 3.1, Proof

Proof of Theorem 3.1

For simplicity, we assume that \Vert h(z)\Vert_\infty\leq 1. In view of lemma 2.2, we have |a_n|\leq 1-|a_0|^2 for n\geq 1. As f is locally univalent and K-quasiconformal sense-preserving harmonic mapping on \mathbb{D}, Schwarz’s Lemma gives that the dilatation \omega=g'/h' is analytic in \mathbb{D} and |\omega(z)|=|g'(z)/h'(z)|\leq k in \mathbb{D}, where K = (1+k)/(1-k) \geq 1, k\in[0,1). Let |a_0|=a\in[0,1]. In view of Lemma 2.4, we have

\begin{aligned} \sum_{n=1}^\infty |b_n|^2 \psi_n(r)\leq k^2 \sum_{n=1}^\infty |a_n|^2 \psi_n(r)\leq k^2\left(1-a^2\right)^2\sum_{n=1}^\infty \psi_n(r).\end{aligned}

Using the Cauchy-Schwarz inequality and the inequality (Proof), we have

\begin{aligned} &&\sum_{n=1}^\infty |b_n| \psi_n(r)\leq \left(\sum_{n=1}^\infty |b_n|^2 \psi_n(r)\right)^{1/2}\left(\sum_{n=1}^\infty \psi_n(r)\right)^{1/2} \leq k(1-a^2)\sum_{n=1}^\infty \psi_n(r)\\[2mm] &&\sum_{n=1}^\infty n|a_n|^2 \left(\psi_n(r)\right)^{2}\leq \left(1-a^2\right)^2\sum_{n=1}^\infty n \left(\psi_n(r)\right)^{2}.\end{aligned}

Let F(x)=(1-x^p)/(1-x^2), where x\in[0,1) and p\in(0,2]. It is evident that

\begin{aligned} F'(x)=\frac{x F_1(x)}{(1-x^2)^2},\quad\text{where}\quad F_1(x)=2 - (2-p)x^p-px^{p-2}.\end{aligned}

Therefore, F_1'(x)=(2-p)px^{p-3} (1-x^2)\geq 0 for x\in[0,1) and p\in(0,2], which shows that F_1(x) is a monotonically increasing function of x. Thus, we have F_1(x)\leq \lim_{x\to1^-} F_1(x)=0 and it’s shows that F(x) is a monotonically decreasing function of x. Hence, we have F(x)\geq \lim_{x\to1^-} F(x)=p/2. Therefore,

\begin{aligned} &&|a_0|^p\psi_0(r)+\sum_{n=1}^\infty |a_n| \psi_n(r)+\sum_{n=1}^\infty |b_n| \psi_n(r)+G\left(\sum_{n=1}^\infty n|a_n|^2(\psi_n(r))^2\right)\\[2mm] &\leq& a^p\psi_0(r)+(1+k)\left(1-a^2\right)\sum_{n=1}^\infty \psi_n(r)+\sum_{t=1}^N c_t\left(1-a^2\right)^{2t}\left(\sum_{n=1}^\infty n \left(\psi_n(r)\right)^2\right)^t\\[2mm] &=\psi_0(r)+\left(1-a^2\right)\left((1+k)\sum_{n=1}^\infty \psi_n(r)+\sum_{t=1}^N c_t \left(1-a^2\right)^{2t-1}\left(\sum_{n=1}^\infty n \left(\psi_n(r)\right)^2\right)^t\right.\\ &&\left.-\left(\frac{1-a^p}{1-a^2}\right)\psi_0(r)\right)\\[2mm] &\leq& \psi_0(r)+\left(1-a^2\right)\left((1+k)\sum_{n=1}^\infty \psi_n(r)+\sum_{t=1}^N c_t \left(1-a^2\right)^{2t-1}\left(\sum_{n=1}^\infty n \left(\psi_n(r)\right)^2\right)^t\right.\\[2mm] &&\left.-\frac{p}{2}\psi_0(r)\right)\\[2mm] &= \psi_0(r)+\left(1-a^2\right)F_1(a, r),\end{aligned}

where k=(K-1)/(K+1) and

\begin{aligned} F_1(a, r)&=(1+k)\sum_{n=1}^\infty \psi_n(r)+\sum_{t=1}^N c_t \left(1-a^2\right)^{2t-1}\left(\sum_{n=1}^\infty n \left(\psi_n(r)\right)^2\right)^t-\frac{p}{2}\psi_0(r).\end{aligned}

Differentiating partially F_1(a, r) with respect to a, we obtain

\begin{aligned} \frac{\partial}{\partial a} F_1(a, r)=-2a\sum_{t=1}^N c_t (2t-1)\left(1-a^2\right)^{2t-2}\left(\sum_{n=1}^\infty n \left(\psi_n(r)\right)^2\right)^t\leq 0.\end{aligned}

Therefore, F_1(a, r) is a monotonically decreasing function of a\in[0,1] and it follows that

\begin{aligned} F_1(a, r)\leq F_1(0,r)=\frac{2K}{K-1}\sum_{n=1}^\infty \psi_n(r)+\sum_{t=1}^N c_t\left(\sum_{n=1}^\infty n \left(\psi_n(r)\right)^2\right)^t-\frac{p}{2}\psi_0(r)=\Phi_1(r)\leq 0\end{aligned}

for r\leq R_{N,K}, under the assumption that (3.1) holds.
To prove the sharpness of the result, we consider the function f_1(z)=h_1(z)+\overline{g_1(z)} in \mathbb{D} such that

\begin{aligned} h_1(z)=\frac{a-z}{1-az}=A_0+\sum_{n=1}^\infty A_n z^n,\end{aligned}

where A_0=a, A_n=-(1-a^2)a^{n-1} for n\geq 1, a\in[0,1) and g_1(z)=\lambda k \sum_{n=1}^\infty A_n z^n, where |\lambda|=1 and k=(K-1)/(K+1). Thus,

\begin{aligned} &&|A_0|^p\psi_0(r)+\sum_{n=1}^\infty |A_n| \psi_n(r)+\sum_{n=1}^\infty |k\lambda A_n| \psi_n(r)+G\left(\sum_{n=1}^\infty n|A_n|^2(\psi_n(r))^2\right)\\[2mm] &=\psi_0(r)+\left(1-a^2\right)(1+k)\sum_{n=1}^\infty a^{n-1} \psi_n(r)\\ &&+\sum_{t=1}^N c_t \left(1-a^2\right)^{2t}\left(\sum_{n=1}^\infty n a^{2(n-1)} \left(\psi_n(r)\right)^2\right)^t-(1-a^p)\psi_0(r)\\[2mm] &= \psi_0(r)+\left(1-a^2\right)F_2(a,r)+\sum_{t=1}^N c_t \left(1-a^2\right)^{2t}\left(\sum_{n=1}^\infty n a^{2(n-1)}\left(\psi_n(r)\right)^2\right)^t\\[2mm] &&-\left(1-a^2\right)\sum_{t=1}^N c_t \left(\sum_{n=1}^\infty n a^{2(n-1)}\left(\psi_n(r)\right)^2\right)^t,\end{aligned}

where k=(K-1)/(K+1) and

\begin{aligned} F_2(a,r)=(1+k)\sum_{n=1}^\infty \psi_n(r)+\sum_{k=1}^N c_k\left(\sum_{n=1}^\infty n a^{2(n-1)}\left(\psi_n(r)\right)^2\right)^k-\frac{1-a^p}{1-a^2}\psi_0(r).\end{aligned}

Letting a\to 1^-, we have

Since \Phi_1(r)>0 for (R_{N,K}, R_{N,K}+\varepsilon), thus the radius cannot be improved. This completes the proof.

Uses Lemma 2.2, Lemma 2.4

Remark

Remark 3.1

PTR0

Setting K=1 and G(x)\equiv 0 in Theorem 3.1 gives Theorem F.

Uses Theorem 3.1

Corollary

Corollary 3.1

9DVV

\begin{aligned} \Phi_2(r):=\frac{2K}{K+1}\sum_{n=1}^\infty \psi_n(r)-\frac{p}{2}\psi_0(r)\leq 0\end{aligned}

for r\leq R_{K}, then the following inequality holds

\begin{aligned} |a_0|^p\psi_0(r)+\sum_{n=1}^\infty |a_n| \psi_n(r)+\sum_{n=1}^\infty |b_n| \psi_n(r)\leq \psi_0(r)\Vert h(z)\Vert_\infty\end{aligned}

for r\leq R_{K}, where R_{K}\in(0,1) is minimal positive root of the equation \Phi_2(r)=0. In the case when \Phi_2(r)>0 in some interval (R_{K}, R_{K}+\varepsilon) (\varepsilon>0), then the number R_{K} cannot be improved.

Used by Proof

Corollary

Corollary 3.2

3TJJ

\begin{aligned} \sum_{n=0}^\infty |a_n|r^n+\sum_{n=1}^\infty |b_n| r^n+c_1\left(\frac{S_r(h)}{\pi}\right)\leq \Vert h\Vert_\infty\quad\text{for}\quad r\leq R_{1,K},\end{aligned}

where R_{1,K}\in(0,1) is the unique positive root of the equation

\begin{aligned} F_3(r):=\frac{2K}{K+1}\left(\frac{r}{1-r}\right)+c_1\frac{r^2}{\left(1-r^2\right)^2}-\frac{1}{2}=0.\end{aligned}

The number R_{1,K} is the best possible.

Remark

Remark 3.2

18N9

Note that, the function F_3(r) is a monotonically increasing function of r\in[0,1] with F_3(0)=-1/2<0 and \lim_{r\to1^-}F_3(r)=+\infty. Therefore, R_{1,K} is the unique positive root of the equation in (0,1).

Corollary

Corollary 3.3

3R2H

\begin{aligned} |a_0|^2+\sum_{n=1}^\infty |a_n|r^n+\sum_{n=1}^\infty |b_n| r^n+c_1\left(\frac{S_r(h)}{\pi}\right)\leq \Vert h\Vert_\infty\quad\text{for}\quad r\leq R_{2,K},\end{aligned}

where R_{2,K}\in(0,1) is the unique positive root of the equation

\begin{aligned} \frac{2K}{K+1}\left(\frac{r}{1-r}\right)+c_1\frac{r^2}{\left(1-r^2\right)^2}-1=0.\end{aligned}

The number R_{2,K} is the best possible.

Corollary

Corollary 3.4

HRGT

\begin{aligned} \sum_{n=0}^\infty |a_n|r^n+\sum_{n=1}^\infty |b_n| r^n+c_1\left(\frac{S_r(h)}{\pi}\right)+c_2\left(\frac{S_r(h)}{\pi}\right)^2+\ldots+c_N\left(\frac{S_r(h)}{\pi}\right)^N\leq \Vert h\Vert_\infty\end{aligned}

for r\leq R_{N,K}, where R_{N,K}\in(0,1) is the unique positive root of the equation

\begin{aligned} F_4(r):=\frac{2K}{K+1}\left(\frac{r}{1-r}\right)+\sum_{t=1}^N c_t\left(\frac{r^2}{\left(1-r^2\right)^2}\right)^t-\frac{1}{2}=0.\end{aligned}

The number R_{N,K} is the best possible.

Remark

Remark 3.3

89NK

It is evident that F_4(r) is a monotonically increasing function of r\in[0,1] with F_4(0)=-1/2<0 and \lim_{r\to1^-}F_4(r)=+\infty.

Corollary

Corollary 3.5

4V3N

\begin{aligned} |a_0|^2+\sum_{n=1}^\infty\left( |a_n|+ |b_n|\right) r^n+\sum_{t=1}^N c_t\left(\frac{S_r(h)}{\pi}\right)^t\leq \Vert h\Vert_\infty\end{aligned}

for r\leq R, where R\in(0,1) is the unique positive root of the equation

\begin{aligned} \frac{2K}{K+1}\left(\frac{r}{1-r}\right)+\sum_{t=1}^N c_t\left(\frac{r^2}{\left(1-r^2\right)^2}\right)^t-1=0.\end{aligned}

Definition

Definition 3.1

QRNY

[SS1989] Let \psi_1 and \psi_2 be two analytic functions in \mathbb{D} given by \psi_1(z)=\sum_{n=0}^{\infty}a_nz^n and \psi_2(z)=\sum_{n=0}^{\infty}b_nz^n. The convolution (or, Hadamard product) is defined by

\begin{aligned}\left(\psi_1*\psi_2\right)(z)=\sum_{n=0}^{\infty}a_nb_nz^n\quad\text{for}\quad z\in\mathbb{D}.\end{aligned}

Definition

Definition 3.2

MSTJ

For parameters a, b, c\in{\Bbb{C}} with c\not=-m (m\in{\Bbb{N}}\cup\{0\}) and z\in{\Bbb{D}}, the Gaussian hypergeometric function is defined by

\begin{aligned} F(a, b, c, z):={}_2F_1(a,b;c;z)=\sum_{n=0}^\infty \gamma_n z^n,\quad\text{where}\quad \gamma_n=\frac{(a)_n(b)_n}{(c)_n(1)_n},\end{aligned}

which converges for all z\in{\Bbb{D}} and converges on the circle |z|=1 if \text{Re}(c-a-b)>0. Here (a)_n is the Pochhammer symbol, which is defined by:

\begin{aligned} (a)_n:=\left\{\begin{array}{lll} 1,&\text{if}\quad n=0\\ a(a+1)(a+2)\cdots(a+n-1),&\text{if}\quad n>0.\end{array}\right. \end{aligned}

Theorem

Theorem 3.2

XUAK

Suppose that f(z)=h(z)+\overline{g(z)}=\sum_{n=0}^\infty a_n z^n+\overline{\sum_{n=1}^\infty b_n z^n} is a sense-preserving K-quasiconformal harmonic mapping in \mathbb{D}, where h(z) is bounded in \mathbb{D} and p\in(0,2]. Let a, b, c\in{\Bbb{R}}^+ be such that (a+n)(b+n)r-(c+n)(1+n)\leq0 for n\geq 0 and r\in[0,1]. Then,

\begin{aligned} |a_0|^p+\sum_{n=1}^\infty \gamma_n|a_n|r^n+\sum_{n=1}^\infty \gamma_n|b_n|r^n\leq \Vert h\Vert_\infty \quad\text{for}\quad r\leq r_0, \end{aligned}

where r_0\in(0,1) is the minimal positive root of the equation F(a, b, c, r)-1=p(K+1)/(4K). The number r_0 cannot be improved.

Proof of Theorem 3.2

Let \psi_n(r)=\gamma_n r^n for n\geq 0. It is evident that \gamma_0=1 and

\begin{aligned} F(a, b, c, r)-1=\sum_{n=1}^\infty \gamma_n r^n=\sum_{n=1}^\infty\psi_n.\end{aligned}

Note that (a)_{n+1}=a(a+1)(a+2)\cdots(a+n)=(a+n)(a)_n for n\in{\Bbb{N}}. Clearly, \gamma_n=(a)_n(b)_n/(c)_n(1)_n>0 for n\geq 1, since a,b,c\in{\Bbb{R}}^+. Therefore, we have

\begin{aligned} &&\gamma_{n+1}=\frac{(a+n)(b+n)}{(c+n)(1+n)}\gamma_n\quad\text{and}\\[2mm] &&\psi_{n+1}(r)-\psi_n(r)=\gamma_{n+1} r^{n+1}-\gamma_n r^n=\gamma_n r^n\left(\frac{(a+n)(b+n)}{(c+n)(1+n)}r-1\right)\leq 0\end{aligned}

for n\geq 0, by the assumption. Thus, \{\psi_n(r)\}_{n=1}^\infty\in{\mathcal{F}} is a decreasing sequence. In view of Corollary 3.1, we have

\begin{aligned} |a_0|^p+\sum_{n=1}^\infty |\gamma_n||a_n|r^n+\sum_{n=1}^\infty |\gamma_n||b_n|r^n\leq \Vert h\Vert_\infty\end{aligned}

holds for r\leq r_0, where r_0 is the minimal positive root of the equation F(a, b, c, r)-1=p(K+1)/(4K).

Uses Corollary 3.1

Theorem

Theorem 4.1

FE1X

\begin{aligned} \Phi_3(r):=\frac{2K}{K+1}\sum_{n=1}^\infty \psi_n(r)+\sum_{t=1}^N c_t\left(\sum_{n=1}^\infty n \left(\psi_n(r)\right)^2\right)^t-\frac{p}{2}\psi_0(r)\leq 0\end{aligned}

for r\leq R_{N,K}, then the following inequality holds

for r\leq R_{N,K}, where \Psi_t(r)=\sum_{k=t}^\infty \psi_k(r), G(x) is defined in (?) and R_{N,K}\in(0,1) is minimal positive root of the equation \Phi_3(r)=0. In the case when \Phi_3(r)>0 in some interval (R_{N,K}, R_{N,K}+\varepsilon) (\varepsilon>0), then the number R_{N,K} cannot be improved.

Used by Remark 4.1

Proof of Theorem 4.1

\begin{aligned} \sum_{n=1}^\infty |b_n|^2 \psi_n(r)\leq k^2 \sum_{n=1}^\infty |a_n|^2 \psi_n(r)\leq k^2\left(1-a^2\right)^2\sum_{n=1}^\infty \psi_n(r),\end{aligned}

where k= (K-1)/(K+1), k\in[0,1) and |a_0|=a\in[0,1). In view of Cauchy-Schwarz inequality and the inequality (Proof), we have

\begin{aligned}\sum_{n=1}^\infty |b_n| \psi_n(r)\leq \left(\sum_{n=1}^\infty |b_n|^2 \psi_n(r)\right)^{1/2}\left(\sum_{n=1}^\infty \psi_n(r)\right)^{1/2} \leq k(1-a^2)\sum_{n=1}^\infty \psi_n(r)\\[2mm]\text{and}\quad \sum_{n=1}^\infty n|a_n|^2 \left(\psi_n(r)\right)^{2}\leq \left(1-a^2\right)^2\sum_{n=1}^\infty n \left(\psi_n(r)\right)^{2}.\hspace{4cm}\end{aligned}

By employing the conclusion of Lemma 2.6, we have

\begin{aligned} \sum_{n=1}^\infty |a_n|\psi_n(r)+\sum_{n=1}^\infty |a_n|^2\left(\frac{\psi_{2n}(r)}{1+|a_0|}+\Psi_{2n+1}(r)\right)\leq \left(1-a^2\right)\Psi_1(r),\end{aligned}

where \Psi_t(r)=\sum_{k=t}^\infty \psi_k(r). From (Proof), (Proof) and (Proof), we have

\begin{aligned} &&|a_0|^p\psi_0(r)+\sum_{n=1}^\infty |a_n| \psi_n(r)+\sum_{n=1}^\infty |b_n| \psi_n(r)+\sum_{n=1}^\infty |a_n|^2\left(\frac{\psi_{2n}(r)}{1+|a_0|}+\Psi_{2n+1}(r)\right)\\ &&+G\left(\sum_{n=1}^\infty n|a_n|^2(\psi_n(r))^2\right)\\[2mm] &\leq& a^p\psi_0(r)+(1+k)\left(1-a^2\right)\sum_{n=1}^\infty \psi_n(r)+\sum_{t=1}^N c_t\left(1-a^2\right)^{2t}\left(\sum_{n=1}^\infty n \left(\psi_n(r)\right)^2\right)^t.\end{aligned}

The rest of the calculations are derived from Theorem 3.1.
To prove the sharpness of the result, we consider the function f_3(z)=h_3(z)+\overline{g_3(z)} in \mathbb{D} such that

\begin{aligned} h_3(z)=\frac{a-z}{1-az}=A_0+\sum_{n=1}^\infty A_n z^n,\end{aligned}

where A_0=a, A_n=-(1-a^2)a^{n-1} for n\geq 1, a\in[0,1) and g_3(z)=\lambda k \sum_{n=1}^\infty A_n z^n, where |\lambda|=1 and k=(K-1)/(K+1). Thus,

\begin{aligned} &&|A_0|^p\psi_0(r)+\sum_{n=1}^\infty |A_n| \psi_n(r)+\sum_{n=1}^\infty |k\lambda A_n| \psi_n(r)\\ &&+\sum_{n=1}^\infty |A_n|^2\left(\frac{\psi_{2n}(r)}{1+|A_0|}+\Psi_{2n+1}(r)\right)+G\left(\sum_{n=1}^\infty n|A_n|^2(\psi_n(r))^2\right)\\[2mm] &= a^p\psi_0(r)+\left(1-a^2\right)(1+k)\sum_{n=1}^\infty a^{n-1} \psi_n(r)+G\left(\left(1-a^2\right)^2\sum_{n=1}^\infty n a^{2n-2}(\psi_n(r))^2\right)\\ &&+\left(1-a^2\right)^2\sum_{n=1}^\infty a^{2n-2}\left(\frac{\psi_{2n}(r)}{1+a}+\Psi_{2n+1}(r)\right)\\[2mm] &= \psi_0(r)+\left(1-a^2\right)F_3(a,r)+\sum_{t=1}^N c_t \left(1-a^2\right)^{2t}\left(\sum_{n=1}^\infty n a^{2n-2}\left(\psi_n(r)\right)^2\right)^t\\[2mm] &&+\left(1-a^2\right)^2\sum_{n=1}^\infty a^{2n-2}\left(\frac{\psi_{2n}(r)}{1+a}+\Psi_{2n+1}(r)\right)\\ &&-\left(1-a^2\right)\sum_{t=1}^N c_t \left(\sum_{n=1}^\infty n a^{2n-2}\left(\psi_n(r)\right)^2\right)^t,\end{aligned}

where k=(K-1)/(K+1) and

\begin{aligned} F_3(a,r)=(1+k)\sum_{n=1}^\infty \psi_n(r)+\sum_{k=1}^N c_k\left(\sum_{n=1}^\infty n a^{2n-2}\left(\psi_n(r)\right)^2\right)^k-\frac{1-a^p}{1-a^2}\psi_0(r).\end{aligned}

Letting a\to 1^-, we have

\begin{aligned}&& |A_0|^p\psi_0(r)+\sum_{n=1}^\infty |A_n| \psi_n(r)+\sum_{n=1}^\infty |k\lambda A_n| \psi_n(r)+\sum_{n=1}^\infty |A_n|^2\left(\frac{\psi_{2n}(r)}{1+|A_0|}+\Psi_{2n+1}(r)\right)\\ &&+G\left(\sum_{n=1}^\infty n|A_n|^2(\psi_n(r))^2\right)\\[2mm] &=\psi_0(r)+\left(1-a^2\right)\Phi_3(r)+O\left(\left(1-a^2\right)^2\right).\end{aligned}

Since \Phi_3(r)>0 for (R_{N,K}, R_{N,K}+\varepsilon), thus the radius cannot be improved. This completes the proof.

Uses Lemma 2.2, Lemma 2.4, Lemma 2.6, Theorem 3.1

Remark

Remark 4.1

6D1C

Setting K=1 and G(x)\equiv 0 in Theorem 4.1 gives Theorem G.

Uses Theorem 4.1

Corollary

Corollary 4.1

V2EV

\begin{aligned} \Phi_4(r):=\frac{2K}{K+1}\sum_{n=1}^\infty \psi_n(r)-\frac{p}{2}\psi_0(r)\leq 0\end{aligned}

for r\leq R_{K}, then the following inequality holds

\begin{aligned} |a_0|^p\psi_0(r)+\sum_{n=1}^\infty (|a_n|+ |b_n|) \psi_n(r)+\sum_{n=1}^\infty |a_n|^2\left(\frac{\psi_{2n}(r)}{1+|a_0|}+\Psi_{2n+1}(r)\right)\leq \psi_0(r)\Vert h(z)\Vert_\infty\end{aligned}

for r\leq R_{K}, where R_{K} is minimal positive root of the equation \Phi_4(r)=0. In the case when \Phi_4(r)>0 in some interval (R_{K}, R_{K}+\varepsilon) (\varepsilon>0), then the number R_{K} cannot be improved.

Theorem

Theorem 4.2

CSMZ

\Phi_5(r):=2\psi_1(r)+2(1+r)^2\left(\frac{2K}{K+1}\sum_{n=2}^\infty \psi_n(r)+\left(\frac{K-1}{K+1}\right)\psi_1(r)\right)-(1-r^2)\psi_0(r)\leq 0

for r\leq R_{K}, then the following inequality holds

\begin{aligned} &&|h(z)|\psi_0(r)+|h'(z)|\psi_1(r)+\sum_{n=2}^\infty |a_n| \psi_n(r)+\sum_{n=1}^\infty |b_n| \psi_n(r)\\[2mm] &&+\sum_{n=1}^\infty |a_n|^2\left(\frac{\psi_{2n}(r)}{1+|a_0|}+\Psi_{2n+1}(r)\right)\leq \psi_0(r)\Vert h(z)\Vert_\infty\end{aligned}

for r\leq R_{K}\leq R, where R\in(0,1) is the minimal positive root of the equation 2\psi_1(r)=\left(1-r^2\right)\psi_0(r) and R_{K} is minimal positive root of the equation \Phi_5(r)=0. In the case when \Phi_5(r)>0 in some interval (R_{K}, R_{K}+\varepsilon) (\varepsilon>0), then the number R_{K} cannot be improved.

Proof of Theorem 4.2

For simplicity, we assume that \Vert h(z)\Vert_\infty\leq 1. In view of lemma 2.2, we have |a_n|\leq 1-|a_0|^2 for n\geq 1. Using arguments as in the proof of Theorem 3.1 and in view of Lemmas 2.2, 2.4 with the condition |g'(z)|\leq k|h'(z)| in {\Bbb{D}}, we have the inequalities (Proof) and (Proof), where k= (K-1)/(K+1), k\in[0,1) and |a_0|=a\in[0,1]. In view of Lemma 2.6, we have

\sum_{n=2}^\infty |a_n|\psi_n(r)+\sum_{n=1}^\infty |a_n|^2\left(\frac{\psi_{2n}(r)}{1+|a_0|}+\Psi_{2n+1}(r)\right)\leq \left(1-a^2\right)\Psi_2(r).

Let Q(x)=x+\beta(1-x^2), where 0\leq x\leq \alpha(\leq1) and \beta\geq 0. Therefore, Q'(x)=1-2\beta x, Q''(x)=-2\beta\leq 0. It follows that Q'(x) is a monotonically decreasing function of x, and hence, we have Q'(x)\geq Q'(\alpha)=1-2\beta\alpha\geq 0 for \beta\leq 1/(2\alpha). Therefore, we have Q(x)\leq Q(\alpha) for 0\leq \beta\leq 1/(2\alpha). In consideration of Lemmas 2.1 and 2.2, and utilizing the inequalities (Proof), (Proof), and (Proof), we have

\begin{aligned}&\leq&\psi_0(r)\left(|h(z)|+\frac{\psi_1(r)}{\left(1-r^2\right)\psi_0(r)}\left(1-|h(z)|^2\right)\right)+(1+k)\left(1-a^2\right)\sum_{n=2}^\infty \psi_n(r)\\[2mm] &&+k\left(1-a^2\right)\psi_1(r)\\[2mm] &\leq&\psi_0(r)\left(\frac{a+r}{1+ar}+\frac{\psi_1(r)}{\left(1-r^2\right)\psi_0(r)}\left(1-\left(\frac{a+r}{1+ar}\right)^2\right)\right)\\[2mm] &&+(1+k)\left(1-a^2\right)\sum_{n=2}^\infty \psi_n(r)+k\left(1-a^2\right)\psi_1(r)\\[2mm] &=\psi_0(r)+\frac{(1-a)F_4(a,r)}{1+ar},\end{aligned}

where

\begin{aligned} F_4(a,r)=\frac{(1+a)\psi_1(r)}{1+ar}+(1+ar)\left(1+a\right)\left((1+k)\sum_{n=2}^\infty \psi_n(r)+k\psi_1(r)\right)-(1-r)\psi_0(r)\end{aligned}

and the inequality (Proof) holds for \psi_1(r)/\left(\left(1-r^2\right)\psi_0(r)\right)\leq 1/2, i.e., for r\in[0,R], where R\in(0,1) is the minimal positive root of the equation 2\psi_1(r)-\left(1-r^2\right)\psi_0(r)=0. Differentiating partially F_4(a,r) with respect to a, we obtain

\begin{aligned}\frac{\partial}{\partial a}F_4(a,r)=\left(1+r+2ar\right)\left((1+k)\sum_{n=2}^\infty \psi_n(r)+k\psi_1(r)\right)+\frac{(1-r)\psi_1(r)}{(1+ar)^2}\geq 0.\end{aligned}

Therefore, F_4(a,r) is a monotonically increasing function of a\in[0, 1] and it follows that F_4(a,r)\leq F_4(1,r) and

\begin{aligned} F_4(1,r)=\frac{1}{1+r}\left(2\psi_1(r)+2(1+r)^2\left((1+k)\sum_{n=2}^\infty \psi_n(r)+k\psi_1(r)\right)-\left(1-r^2\right)\psi_0(r)\right),\end{aligned}

which is less than or equal to zero for r\leq R_K, under the assumption that (4.2) holds. We now assert that R_K\leq R. For r>R, we have 2\psi_1(r)>\left(1-r^2\right)\psi_0 and

\begin{aligned}\Phi_5(r)>2(1+r)^2\left(\frac{2K}{K+1}\sum_{n=2}^\infty \psi_n(r)+\left(\frac{K-1}{K+1}\right)\psi_1(r)\right)\geq0,\end{aligned}

which shows that R_K\leq R.
To prove the sharpness of the result, we consider the function f_4(z)=h_4(z)+\overline{g_4(z)} in \mathbb{D} such that

\begin{aligned} h_4(z)=\frac{a-z}{1-az}=A_0+\sum_{n=1}^\infty A_n z^n,\end{aligned}

where A_0=a, A_n=-(1-a^2)a^{n-1} for n\geq 1, a\in[0,1) and g_4(z)=\lambda k \sum_{n=1}^\infty A_n z^n, where |\lambda|=1 and k=(K-1)/(K+1). Also h_4'(z)=-\left(1-a^2\right)/(1-az)^2. Thus,

\begin{aligned} &&|h_4(-r)|\psi_0(r)+|h_4'(-r)|\psi_1(r)+\sum_{n=2}^\infty |A_n| \psi_n(r)+\sum_{n=1}^\infty |\lambda k A_n| \psi_n(r)\\[2mm] &&+\sum_{n=1}^\infty |A_n|^2\left(\frac{\psi_{2n}(r)}{1+|A_0|}+\Psi_{2n+1}(r)\right)\\[2mm] &=\left(\frac{a+r}{1+ar}\right)\psi_0(r)+\left(\frac{1-a^2}{(1+ar)^2}\right)\psi_1(r)+\left(1-a^2\right)(1+k)\sum_{n=2}^\infty a^{n-1} \psi_n(r)\\[2mm] &&+k\left(1-a^2\right)\psi_1(r)+\left(1-a^2\right)^2\sum_{n=1}^\infty a^{2n-2}\left(\frac{\psi_{2n}(r)}{1+a}+\Psi_{2n+1}(r)\right)\\[2mm] &=\psi_0(r)+\frac{(1-a)F_5(a,r)}{(1+ar)^2}+\left(1-a^2\right)^2\sum_{n=1}^\infty a^{2n-2}\left(\frac{\psi_{2n}(r)}{1+a}+\Psi_{2n+1}(r)\right),\end{aligned}

where k=(K-1)/(K+1) and

\begin{aligned} F_5(a,r)&=\left(1+a\right)\psi_1(r)+(1+ar)^2\left(1+a\right)\left((1+k)\sum_{n=2}^\infty \psi_n(r)+k\psi_1(r)\right)\\[2mm] &&-(1-r)(1+ar)\psi_0(r).\end{aligned}

Letting a\to 1^-, we have

\begin{aligned}&&|h_4(-r)|\psi_0(r)+|h_4'(-r)|\psi_1(r)+\sum_{n=2}^\infty |A_n| \psi_n(r)+\sum_{n=1}^\infty |B_n| \psi_n(r)\\[2mm] &&+\sum_{n=1}^\infty |A_n|^2\left(\frac{\psi_{2n}(r)}{1+|A_0|}+\Psi_{2n+1}(r)\right)\hspace{10cm}\\[2mm] &=\psi_0(r)+\frac{(1-a)}{(1+ar)^2}\Phi_5(r)+O\left(\left(1-a\right)^2\right).\end{aligned}

Since \Phi_5(r)>0 for (R_{K}, R_{K}+\varepsilon), thus the radius cannot be improved. This completes the proof.

Uses Lemma 2.2, Lemma 2.4, Proof, Lemma 2.6, Lemma 2.1

Theorem

Theorem 4.3

VJB0

\Phi_6(r):=\psi_1(r)+(1+r)^2\left(\frac{2K}{K+1}\sum_{n=2}^\infty \psi_n(r)+\left(\frac{K-1}{K+1}\right)\psi_1(r)\right)-\left(1-r^2\right)\psi_0(r)\leq 0

for r\leq R_{K}, then the following inequality holds

for r\leq R_{K}\leq R, where R\in(0,1) is the minimal positive root of the equation 2\psi_1(r)=\left(1-r^2\right)\psi_0(r) and R_{K} is minimal positive root of the equation \Phi_6(r)=0. In the case when \Phi_6(r)>0 in some interval (R_{K}, R_{K}+\varepsilon) (\varepsilon>0), then the number R_{K} cannot be improved.

Proof of Theorem 4.3

For simplicity, we assume that \Vert h(z)\Vert_\infty\leq 1. In view of lemma 2.2, we have |a_n|\leq 1-|a_0|^2 for n\geq 1. Let |a_0|=a\in[0,1]. Using arguments as in the proof of Theorem 4.2 and in view of Lemmas 2.2, 2.4 and 2.6 with the condition |g'(z)|\leq k|h'(z)| in {\Bbb{D}}, we have the inequalities (Proof), (Proof), (Proof), where k= (K-1)/(K+1), k\in[0,1). In consideration of Lemmas 2.1 and 2.2, and utilizing the inequalities (Proof), (Proof), and (Proof), we have

\begin{aligned} &&|h(z)|^2\psi_0(r)+|h'(z)|\psi_1(r)+\sum_{n=2}^\infty |a_n| \psi_n(r)+\sum_{n=1}^\infty |b_n| \psi_n(r)\\[2mm] &&+\sum_{n=1}^\infty |a_n|^2\left(\frac{\psi_{2n}(r)}{1+|a_0|}+\Psi_{2n+1}(r)\right)\\[2mm] &\leq&\psi_0(r)\left(|h(z)|^2+\frac{\psi_1(r)}{\left(1-r^2\right)\psi_0(r)}\left(1-|h(z)|^2\right)\right)+(1+k)\left(1-a^2\right)\sum_{n=2}^\infty \psi_n(r)\\[2mm] &&+k\left(1-a^2\right)\psi_1(r)\\[2mm] &\leq&\psi_0(r)\left(\left(1-\frac{\psi_1(r)}{\left(1-r^2\right)\psi_0(r)}\right)\left(\frac{a+r}{1+ar}\right)^2+\frac{\psi_1(r)}{\left(1-r^2\right)\psi_0(r)}\right)\\[2mm] &&+(1+k)\left(1-a^2\right)\sum_{n=2}^\infty \psi_n(r)+k\left(1-a^2\right)\psi_1(r)\\[2mm] &=\psi_0(r)-\left(1-a^2\right)F_6(a,r),\end{aligned}

where

\begin{aligned} F_6(a,r)=\left(\psi_0(r)-\frac{\psi_1(r)}{\left(1-r^2\right)}\right)\frac{\left(1-r^2\right)}{(1+ar)^2}-(1+k)\sum_{n=2}^\infty \psi_n(r)-k\psi_1(r).\end{aligned}

and the inequality (Proof) holds for \psi_1(r)/\left(\left(1-r^2\right)\psi_0(r)\right)\leq 1, i.e., for r\in[0,R], where R\in(0,1) is the minimal positive root of the equation \psi_1(r)-\left(1-r^2\right)\psi_0(r)=0. Differentiating partially F_6(a,r) with respect to a, we obtain

\begin{aligned}\frac{\partial}{\partial a}F_6(a,r)=-\left(\psi_0(r)-\frac{\psi_1(r)}{\left(1-r^2\right)}\right)\frac{2r\left(1-r^2\right)}{(1+ar)^3}\leq 0.\end{aligned}

Therefore, F_6(a,r) is a monotonically decreasing function of a\in[0, 1] and it follows that F_6(a,r)\geq F_6(1,r) and

\begin{aligned} F_6(1,r)=\frac{1}{(1+r)^2}\left(\left(1-r^2\right)\psi_0(r)-(1+r)^2\left((1+k)\sum_{n=2}^\infty \psi_n(r)+k\psi_1(r)\right)-\psi_1(r)\right),\end{aligned}

which is greater than or equal to zero for r\leq R_K, under the assumption that (4.3) holds. We now assert that R_K\leq R. For r>R, we have \psi_1(r)>\left(1-r^2\right)\psi_0 and

\begin{aligned} \Phi_6(r)>(1+r)^2\left(\frac{2K}{K+1}\sum_{n=2}^\infty \psi_n(r)+\left(\frac{K-1}{K+1}\right)\psi_1(r)\right)\geq 0,\end{aligned}

which shows that R_K\leq R.
To prove the sharpness of the result, we consider the function f_5(z)=h_5(z)+\overline{g_5(z)} in \mathbb{D} such that

\begin{aligned} h_5(z)=\frac{a-z}{1-az}=A_0+\sum_{n=1}^\infty A_n z^n,\end{aligned}

where A_0=a, A_n=-(1-a^2)a^{n-1} for n\geq 1, a\in[0,1) and g_4(z)=\lambda k \sum_{n=1}^\infty A_n z^n, where |\lambda|=1 and k=(K-1)/(K+1). Also h_5'(z)=-\left(1-a^2\right)/(1-az)^2. Thus,

\begin{aligned} &&|h_4(-r)|^2\psi_0(r)+|h_4'(-r)|\psi_1(r)+\sum_{n=2}^\infty |A_n| \psi_n(r)+\sum_{n=1}^\infty |\lambda k A_n| \psi_n(r)\\[2mm] &&+\sum_{n=1}^\infty |A_n|^2\left(\frac{\psi_{2n}(r)}{1+|A_0|}+\Psi_{2n+1}(r)\right)\\[2mm] &=\left(\frac{a+r}{1+ar}\right)^2\psi_0(r)+\frac{\left(1-a^2\right)\psi_1(r)}{(1+ar)^2}+\left(1-a^2\right)(1+k)\sum_{n=2}^\infty a^{n-1} \psi_n(r)\\[2mm] &&+k\left(1-a^2\right)\psi_1(r)+\left(1-a^2\right)^2\sum_{n=1}^\infty a^{2n-2}\left(\frac{\psi_{2n}(r)}{1+a}+\Psi_{2n+1}(r)\right)\\[2mm] &=\psi_0(r)+\frac{\left(1-a^2\right)F_7(a,r)}{(1+ar)^2}+\left(1-a^2\right)^2\sum_{n=1}^\infty a^{2n-2}\left(\frac{\psi_{2n}(r)}{1+a}+\Psi_{2n+1}(r)\right),\end{aligned}

where k=(K-1)/(K+1) and

\begin{aligned} F_7(a,r)=\psi_1(r)+(1+ar)^2\left((1+k)\sum_{n=2}^\infty \psi_n(r)+k\psi_1(r)\right)-\left(1-r^2\right)\psi_0(r).\end{aligned}

Letting a\to 1^-, we have

\begin{aligned}&&|h_4(-r)|^2\psi_0(r)+|h_4'(-r)|\psi_1(r)+\sum_{n=2}^\infty |A_n| \psi_n(r)+\sum_{n=1}^\infty |B_n| \psi_n(r)\\[2mm] &&+\sum_{n=1}^\infty |A_n|^2\left(\frac{\psi_{2n}(r)}{1+|A_0|}+\Psi_{2n+1}(r)\right)\hspace{10cm}\\[2mm] &=\psi_0(r)+\frac{\left(1-a^2\right)}{(1+ar)^2}\Phi_6(r)+O\left(\left(1-a^2\right)^2\right).\end{aligned}

Since \Phi_6(r)>0 for (R_{K}, R_{K}+\varepsilon), thus the radius cannot be improved. This completes the proof.

Uses Lemma 2.2, Lemma 2.4, Lemma 2.6, Proof, Proof, Lemma 2.1

Source and licence

The text on this page is from arXiv:2411.01837v1 by Raju Biswas and Rajib Mandal, under the licence CC BY 4.0. The corpus made these changes: it read the items out of the LaTeX source, it expanded the macros of the paper, it put each proof after its statement, and it converted the text to HTML. The formulas are the TeX of the paper, and KaTeX renders them. A formula or a reference can look different from the PDF. The page is published under CC BY 4.0.

The record of each item is in analysis/objects/2411/2411.01837.jsonl.