Next Article in Journal
Steady-State Navier–Stokes Equations in Thin Tube Structure with the Bernoulli Pressure Inflow Boundary Conditions: Asymptotic Analysis
Previous Article in Journal
On Golden Lorentzian Manifolds Equipped with Generalized Symmetric Metric Connection
Previous Article in Special Issue
Modified Operators Interpolating at Endpoints
 
 
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

On an Extension of a Hardy–Hilbert-Type Inequality with Multi-Parameters

1
Department of Mathematics, Guangdong University of Education, Guangzhou 510303, China
2
Department of Mathematics and Engineering Sciences, Hellenic Military Academy, 16673 Vari Attikis, Greece
3
Moscow Institute of Physics and Technology, Institutskiy Per, d. 9, 141700 Dolgoprudny, Russia
4
Program in Interdisciplinary Studies, Institute for Advanced Study, 1 Einstein Dr, Princeton, NJ 08540, USA
5
Department of Mechanics and Mathematics, Moscow State University, 119991 Moscow, Russia
6
Institute for Mathematics and Informatics, Buryat State University, 670000 Ulan-Ude, Russia
7
Caucasus Mathematical Center, Adyghe State University, 385000 Maykop, Russia
*
Author to whom correspondence should be addressed.
Mathematics 2021, 9(19), 2432; https://doi.org/10.3390/math9192432
Received: 9 September 2021 / Revised: 24 September 2021 / Accepted: 27 September 2021 / Published: 30 September 2021
(This article belongs to the Special Issue Mathematical Analysis and Analytic Number Theory 2020)

Abstract

:
Making use of weight coefficients as well as real/complex analytic methods, an extension of a Hardy–Hilbert-type inequality with a best possible constant factor and multiparameters is established. Equivalent forms, reverses, operator expression with the norm, and a few particular cases are also considered.

1. Introduction

In this paper, we generalize the classical Hardy–Hilbert inequality, which can be stated as follows: assume that p > 1 , 1 p + 1 q = 1 , f ( x ) , g ( y ) 0 , f L p ( R + ) , g L q ( R + ) ,
| | f | | p = 0 f p ( x ) d x 1 p > 0 ,
| | g | | q > 0 . We have the following Hardy–Hilbert integral inequality (cf. [1]):
0 0 f ( x ) g ( y ) x + y d x d y < π sin ( π / p ) | | f | | p | | g | | q ,
with the best possible constant factor π sin ( π / p ) .
If a m ,   b n 0 , a = { a m } m = 1 l p , b = { b n } n = 1 l q ,
| | a | | p = m = 1 a m p 1 p > 0 ,
| | b | | q > 0 , then we have the following Hardy–Hilbert inequality with the same best possible constant factor π sin ( π / p ) (cf. [1]):
m = 1 n = 1 a m b n m + n < π sin ( π / p ) | | a | | p | | b | | q .
Inequalities (1) and (2) are important in analysis and its applications (cf. [1,2,3,4,5,6,7,8,9,10,11,12]).
Assuming that μ i , υ j > 0 ( i , j N = { 1 , 2 , } ) ,
U m : = i = 1 m μ i , V n : = j = 1 n υ j ( m , n N ) ,
we have the following inequality (cf. [1], Theorem 321, replacing μ m 1 / q a m (resp. υ n 1 / p b n ) by a m (resp. b n ) ):
m = 1 n = 1 a m b n U m + V n < π sin ( π p ) m = 1 a m p μ m p 1 1 p n = 1 b n q υ n q 1 1 q .
For μ i = υ j = 1 ( i , j N ) , (4) reduces to (2). Inequality (4) is known as Hardy–Hilbert-type inequality.
Note. The authors of [1] did not prove that the constant factor in (4) is the best possible.
In 1998, by introducing an independent parameter λ ( 0 , 1 ] , Yang [13] provided an extension of (1) for p = q = 2 . Improving upon the method of [13], Yang [6] presented the following best possible extensions of (1) and (2):
If λ 1 , λ 2 R , λ 1 + λ 2 = λ , k λ ( x , y ) is a nonnegative homogeneous function of degree λ , with
k ( λ 1 ) = 0 k λ ( t , 1 ) t λ 1 1 d t R + ,
ϕ ( x ) = x p ( 1 λ 1 ) 1 , ψ ( x ) = x q ( 1 λ 2 ) 1 , f ( x ) , g ( y ) 0 ,
f L p , ϕ ( R + ) = f ; | | f | | p , ϕ : = { 0 ϕ ( x ) | f ( x ) | p d x } 1 p < ,
g L q , ψ ( R + ) , | | f | | p , ϕ , | | g | | q , ψ > 0 , then
0 0 k λ ( x , y ) f ( x ) g ( y ) d x d y < k ( λ 1 ) | | f | | p , ϕ | | g | | q , ψ ,
where the constant factor k ( λ 1 ) is the best possible. Moreover, if k λ ( x , y ) keeps a finite value and k λ ( x , y ) x λ 1 1 ( k λ ( x , y ) y λ 2 1 ) is decreasing with respect to x > 0 ( y > 0 ) , then for a m , b n 0 ,
a l p , ϕ = a ; | | a | | p , ϕ : = { n = 1 ϕ ( n ) | a n | p } 1 p < ,
b = { b n } n = 1 l q , ψ , | | a | | p , ϕ , | | b | | q , ψ > 0 , it follows that
m = 1 n = 1 k λ ( m , n ) a m b n < k ( λ 1 ) | | a | | p , ϕ | | b | | q , ψ ,
where the constant factor k ( λ 1 ) is still the best possible.
Clearly, for
λ = 1 , k 1 ( x , y ) = 1 x + y , λ 1 = 1 q , λ 2 = 1 p ,
inequality (5) reduces to (1), while (6) reduces to (2).
For s N , 0 < λ 1 , λ 2 1 , λ 1 + λ 2 = λ , we set
k λ ( x , y ) = 1 k = 1 s ( x λ / s + c k y λ / s ) ( 0 < c 1 < < c s ) .
Then, by (6), we derive that (cf. [14])
m = 1 n = 1 a m b n k = 1 s ( m λ / s + c k n λ / s ) < k s ( λ 1 ) | | a | | p , ϕ | | b | | q , ψ ,
where the constant factor
k s ( λ 1 ) = π s λ sin ( π s λ 1 λ ) k = 1 s c k s λ 1 λ 1 j = 1 ( j k ) s 1 c j c k ( R + )
is the best possible.
Some other kinds of results, such as Hilbert-type integral inequalities, half-discrete Hilbert-type inequalities, and multidimensional Hilbert-type inequalities are provided in [15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42].
In the present paper, making use of weight coefficients as well as real/complex analytic methods, a Hardy–Hilbert-type inequality with a best possible constant factor and multiparameters is established (for p > 1 ). This inequality constitutes an extension of (4) and (7). Equivalent forms, reverses (two cases of 0 < p < 1 and p < 0 ), operator expression with the norm, and a few particular cases are also considered.

2. Some Lemmas

In this section we prove the inequalities of the weight functions, which are used to prove the main results. In the sequel, we assume for the multiparameters that
p ( , 0 ) ( 0 , 1 ) ( 1 , ) ,
1 p + 1 q = 1 , 0 < λ 1 , λ 2 1 , λ 1 + λ 2 = λ , 0 < c 1 c s ( s N ) ,
k s ( λ 1 ) is indicated by (8), μ i , υ j > 0 ( i , j N ) , U m and V n are defined by (3), a m , b n 0 ( m , n N ) ,
| | a | | p , Φ λ = m = 1 Φ λ ( m ) a m p 1 p R + , | | b | | q , Ψ λ = n = 1 Ψ λ ( n ) b n q 1 q R + ,
where we define
Φ λ ( m ) : = U m p ( 1 λ 1 ) 1 μ m p 1 , Ψ λ ( n ) : = V n q ( 1 λ 2 ) 1 υ n q 1 ( m , n N ) .
Lemma 1.
If C is the set of complex numbers and C = C { } ,
z k C \ { z | R e ( z ) 0 , I m ( z ) = 0 } ( k = 1 , 2 , , n )
are different points, the function f ( z ) is analytic in C except for z i ( i = 1 , 2 , , n ) , and z = is a zero point of f ( z ) whose order is not less than 1, then for α R , we have
0 f ( x ) x α 1 d x = 2 π i 1 e 2 π α i k = 1 n R e ( s ) [ f ( z ) z α 1 , z k ] ,
where
0 < I m ( ln z ) = arg z < 2 π .
In particular, if z k ( k = 1 , , n ) are all poles of order 1, setting
φ k ( z ) = ( z z k ) f ( z ) ( φ k ( z k ) 0 ) ,
then
0 f ( x ) x α 1 d x = π sin π α k = 1 n ( z k ) α 1 φ k ( z k ) .
Proof. 
By [43] (p. 118), we obtain (9). We have that
1 e 2 π α i = 1 cos 2 π α i sin 2 π α = 2 i sin π α ( cos π α + i sin π α ) = 2 i e i π α sin π α .
In particular, since
f ( z ) z α 1 = 1 z z k ( φ k ( z ) z α 1 ) ,
it is obvious that
R e ( s ) [ f ( z ) z α 1 , a k ] = z k α 1 φ k ( z k ) = e i π α ( z k ) α 1 φ k ( z k ) .
Then, by (9), we obtain (10).
This completes the proof of the lemma. □
Example 1.
For s N , ε > 0 , we set
k λ ( x , y ) = 1 k = 1 s ( x λ / s + c k y λ / s ) ,
and c ˜ k = c k + ( k 1 ) ε ( k = 1 , , s ) . By (10), we get that
k ˜ s ( λ 1 ) : = 0 k = 1 s 1 t λ / s + c ˜ k t λ 1 1 d t = s λ 0 k = 1 s 1 u + c ˜ k u s λ 1 λ 1 d u = π s λ sin ( π s λ 1 λ ) k = 1 s c ˜ k s λ 1 λ 1 j = 1 ( j k ) s 1 c ˜ j c ˜ k R + .
Since we have
0 < k ˜ s ( λ 1 ) = s λ 0 k = 1 s 1 u + c ˜ k u s λ 1 λ 1 d u s λ 0 1 ( u + c 1 ) s u s λ 1 λ 1 d u = s λ c 1 ( s λ 2 ) / λ 0 1 ( v + 1 ) s v s λ 1 λ 1 d v = s λ c 1 ( s λ 2 ) / λ B ( s λ 1 λ , s λ 2 λ ) R + ,
it follows that
k s ( λ 1 ) = lim ε 0 + k ˜ s ( λ 1 ) = π s λ sin ( π s λ 1 λ ) k = 1 s c k s λ 1 λ 1 j = 1 ( j k ) s 1 c j c k R + .
In particular, for s = 1 , we obtain
k 1 ( λ 1 ) = 1 λ 0 u ( λ 1 / λ ) 1 u + c 1 d u = π λ c 1 λ 2 / λ sin ( π λ 1 λ ) ;
for c s = = c 1 , we derive that
k ( λ 1 ) : = 0 t λ 1 1 ( t λ / s + c 1 ) s d t = s λ c 1 ( s λ 2 ) / λ B ( s λ 1 λ , s λ 2 λ ) .
Lemma 2.
Define the following weight coefficients:
ω s ( λ 2 , m ) : = n = 1 1 k = 1 s ( U m λ / s + c k V n λ / s ) U m λ 1 υ n V n 1 λ 2 , m N ,
ϖ s ( λ 1 , n ) : = m = 1 1 k = 1 s ( U m λ / s + c k V n λ / s ) V n λ 2 μ m U m 1 λ 1 , n N .
Then, we have the following inequalities:
ω s ( λ 2 , m ) < k s ( λ 1 ) ( 0 < λ 2 1 , λ 1 > 0 ; m N ) ,
ϖ s ( λ 1 , n ) < k s ( λ 1 ) ( 0 < λ 1 1 , λ 2 > 0 ; n N ) .
Proof. 
We set
μ ( t ) : = μ m , t ( m 1 , m ] ( m N ) ; υ ( t ) : = υ n , t ( n 1 , n ] ( n N ) ,
U ( x ) : = 0 x μ ( t ) d t ( x 0 ) , V ( y ) : = 0 y υ ( t ) d t ( y 0 ) .
Then by (3), it follows that
U ( m ) = U m , V ( n ) = V n ( m , n N ) .
For x ( m 1 , m ] ,
U ( x ) = μ ( x ) = μ m ( m N ) ;
for y ( n 1 , n ] ,
V ( y ) = υ ( y ) = υ n ( n N ) .
Since V ( y ) is strictly increasing in ( n 1 , n ] , λ s > 0 and 1 λ 2 0 , in view of the decreasing property, we obtain that
ω s ( λ 2 , m ) = n = 1 n 1 n 1 k = 1 s ( U m λ / s + c k V n λ / s ) U m λ 1 V n 1 λ 2 V ( y ) d y < n = 1 n 1 n 1 k = 1 s ( U m λ / s + c k V λ / s ( y ) ) U m λ 1 V 1 λ 2 ( y ) V ( y ) d y .
Setting
t = U m V ( y ) λ / s ,
we obtain
V ( y ) d y = s λ U m t s λ 1 d t
and
ω s ( λ 2 , m ) < s λ n = 1 ( U m V ( n 1 ) ) λ / s ( U m V ( n ) ) λ / s 1 k = 1 s ( t + c k ) t s λ 1 λ 1 d t = s λ ( U m V ( ) ) λ / s 1 k = 1 s ( t + c k ) t s λ 1 λ 1 d t s λ 0 1 k = 1 s ( t + c k ) t s λ 1 λ 1 d t = k s ( λ 1 ) .
Since U ( x ) is strictly increasing in ( m 1 , m ] , λ s > 0 and 0 < λ 1 1 , similarly, we have
ϖ s ( λ 1 , n ) = m = 1 m 1 m 1 k = 1 s ( U m λ / s + c k V n λ / s ) V n λ 2 U ( x ) U m 1 λ 1 d x < m = 1 m 1 m 1 k = 1 s ( U λ / s ( x ) + c k V n λ / s ) V n λ 2 U ( x ) U 1 λ 1 ( x ) d x ( t = ( U ( x ) V n ) λ / s ) = s λ m = 1 ( U ( m 1 ) V ( n ) ) λ / s ( U ( m ) V ( n ) ) λ / s t s λ 1 λ 1 k = 1 s ( t + c k ) d t = s λ 0 ( U ( ) V ( n ) ) λ / s 1 k = 1 s ( t + c k ) t s λ 1 λ 1 d t k s ( λ 1 ) .
Hence, we deduce (15) and (16).
This completes the proof of the lemma. □
Lemma 3.
If m 0 , n 0 N , μ m μ m + 1 ( m { m 0 , m 0 + 1 , } ) , υ n υ n + 1 ( n { n 0 , n 0 + 1 , } ) , U ( ) = V ( ) = , then
(i) for m , n N , we have
k s ( λ 1 ) ( 1 θ ( λ 2 , m ) ) < ω s ( λ 2 , m ) ( 0 < λ 2 1 , λ 1 > 0 ) ,
k s ( λ 1 ) ( 1 ϑ ( λ 1 , n ) ) < ϖ s ( λ 1 , n ) ( 0 < λ 1 1 , λ 2 > 0 ) ,
where
θ ( λ 2 , m ) = O 1 U m λ 2 ( 0 , 1 ) , ϑ ( λ 1 , n ) = O 1 V n λ 1 ( 0 , 1 ) ;
(ii) for any a > 0 , we have
m = 1 μ m U m 1 + a = 1 a 1 U m 0 a + a O ( 1 ) ,
n = 1 υ n V n 1 + a = 1 a 1 V n 0 a + a O ˜ ( 1 ) .
Proof. 
Since υ n υ n + 1 ( n n 0 ) , 1 λ 2 0 and V ( ) = , we have
ω s ( λ 2 , m ) n = n 0 1 k = 1 s ( U m λ / s + c k V n λ / s ) U m λ 1 V n 1 λ 2 υ n + 1 > n = n 0 n n + 1 1 k = 1 s ( U m λ / s + c k V λ / s ( y ) ) U m λ 1 V 1 λ 2 ( y ) V ( y ) d y = s λ n = n 0 ( U m V ( n ) ) λ / s ( U m V ( n + 1 ) ) λ / s 1 k = 1 s ( t + c k ) t s λ 1 λ 1 d t = s λ ( U m V ( ) ) λ / s ( U m V n 0 ) λ / s 1 k = 1 s ( t + c k ) t s λ 1 λ 1 d t = s λ 0 ( U m V n 0 ) λ / s t s λ 1 λ 1 k = 1 s ( t + c k ) d t = k s ( λ 1 ) ( 1 θ ( λ 2 , m ) ) ,
where
θ ( λ 2 , m ) : = s λ k s ( λ 1 ) ( U m V n 0 ) λ / s 1 k = 1 s ( t + c k ) t s λ 1 λ 1 d t ( 0 , 1 ) .
We obtain
0 < θ ( λ 2 , m ) s λ k s ( λ 1 ) ( U m V n 0 ) λ / s 1 t s t s λ 1 λ 1 d t = s λ k s ( λ 1 ) ( U m V n 0 ) λ / s t s λ 2 λ 1 d t = 1 λ 2 k s ( λ 1 ) ( V n 0 U m ) λ 2 ,
and then
θ ( λ 2 , m ) = O 1 U m λ 2 .
Hence, we deduce (17). Similarly, we obtain (18).
For a > 0 , we have that
m = 1 μ m U m 1 + a = m = 1 m 0 μ m U m 1 + a + m = m 0 + 1 μ m U m 1 + a = m = 1 m 0 μ m U m 1 + a + m = m 0 + 1 m 1 m U ( x ) U m 1 + a d x < m = 1 m 0 μ m U m 1 + a + m = m 0 + 1 m 1 m U ( x ) U 1 + a ( x ) d x = m = 1 m 0 μ m U m 1 + a + m 0 d U ( x ) U 1 + a ( x ) = m = 1 m 0 μ m U m 1 + a + 1 a U m 0 a = 1 a 1 U m 0 a + a m = 1 m 0 μ m U m 1 + a ,
m = 1 μ m U m 1 + a m = m 0 μ m + 1 U m 1 + a = m = m 0 m m + 1 U ( x ) U m 1 + a d x
> m = m 0 m m + 1 U ( x ) d x U 1 + a ( x ) = m 0 d U ( x ) U 1 + a ( x ) = 1 a U m 0 a .
Hence, we derive (19). Similarly, we also get (20).
This completes the proof of the lemma. □

3. Main Results and Operator Expressions

In this section, by using Lemma 3, we obtain Theorems 1 and 2.
Theorem 1.
For p > 1 , we have the following equivalent inequalities:
I : = n = 1 m = 1 a m b n k = 1 s ( U m λ s + c k V n λ s ) < k s ( λ 1 ) | | a | | p , Φ λ | | b | | q , Ψ λ ,
J : = n = 1 υ n V n 1 p λ 2 m = 1 a m k = 1 s ( U m λ s + c k V n λ s ) p 1 p < k s ( λ 1 ) | | a | | p , Φ λ .
Proof. 
By Hölder’s inequality with weight (cf. [44]), we have
m = 1 a m k = 1 s ( U m λ / s + c k V n λ / s ) p = m = 1 1 k = 1 s ( U m λ s + c k V n λ s ) U m ( 1 λ 1 ) / q υ n 1 / p a m V n ( 1 λ 2 ) / p μ m 1 / q V n ( 1 λ 2 ) / p μ m 1 / q U m ( 1 λ 1 ) / q υ n 1 / p p m = 1 1 k = 1 s ( U m λ / s + c k V n λ / s ) U m ( 1 λ 1 ) p / q υ n V n 1 λ 2 μ m p / q a m p × m = 1 1 k = 1 s ( U m λ / s + c k V n λ / s ) V n ( 1 λ 2 ) ( q 1 ) μ m U m 1 λ 1 υ n q 1 p 1 = ( ϖ s ( λ 1 , n ) ) p 1 V n p λ 2 1 υ n m = 1 1 k = 1 s ( U m λ / s + c k V n λ / s ) U m ( 1 λ 1 ) ( p 1 ) υ n V n 1 λ 2 μ m p 1 a m p .
In view of (16), we obtain that
J ( k s ( λ 1 ) ) 1 q n = 1 m = 1 1 k = 1 s ( U m λ / s + c k V n λ / s ) U m ( 1 λ 1 ) ( p 1 ) υ n V n 1 λ 2 μ m p 1 a m p 1 p = ( k s ( λ 1 ) ) 1 q m = 1 n = 1 1 k = 1 s ( U m λ / s + c k V n λ / s ) U m ( 1 λ 1 ) ( p 1 ) υ n V n 1 λ 2 μ m p 1 a m p 1 p = ( k s ( λ 1 ) ) 1 q m = 1 ω s ( λ 2 , m ) U m p ( 1 λ 1 ) 1 μ m p 1 a m p 1 p .
Then, by (15), we have (22).
By Hölder’s inequality (cf. [44]), we obtain that
I = n = 1 υ n 1 / p V n 1 p λ 2 m = 1 a m k = 1 s ( U m λ / s + c k V n λ / s ) V n 1 p λ 2 υ n 1 / p b n J | | b | | q , Ψ λ .
Then, by (22), we derive (21). On the other hand, assuming that (21) is valid, we set
b n : = υ n V n 1 p λ 2 m = 1 a m k = 1 s ( U m λ / s + c k V n λ / s ) p 1 , n N .
Then, we get that J p = | | b | | q , Ψ λ q .
If J = 0 , then (22) is trivially valid; if J = , then by (24) and (15), this is impossible. Suppose that 0 < J < . By (21), it follows that
| | b | | q , Ψ λ q = J p = I < k s ( λ 1 ) | | a | | p , Φ λ | | b | | q , Ψ λ ,
| | b | | q , Ψ λ q 1 = J < k s ( λ 1 ) | | a | | p , Φ λ ,
and then (22) follows, which is equivalent to (21).
This completes the proof of the theorem. □
Theorem 2.
If p > 1 , m 0 , n 0 N , μ m μ m + 1 ( m { m 0 , m 0 + 1 , } ) , υ n υ n + 1 ( n { n 0 , n 0 + 1 , } ) , U ( ) = V ( ) = , then the constant factor k s ( λ 1 ) in (21) and (22) is the best possible.
Proof. 
For ε ( 0 , p λ 1 ) , we set
λ ˜ 1 = λ 1 ε p ( ( 0 , 1 ) ) , λ ˜ 2 = λ 2 + ε p ( > 0 ) ,
and
a ˜ m : = U m λ ˜ 1 1 μ m = U m λ 1 ε p 1 μ m , b ˜ n = V n λ ˜ 2 ε 1 υ n = V n λ 2 ε q 1 υ n .
Then, by (19) and (20), we have
| | a ˜ | | p , Φ λ | | b ˜ | | q , Ψ λ = m = 1 μ m U m 1 + ε 1 p n = 1 υ n V n 1 + ε 1 q = 1 ε 1 U m 0 ε + ε O ( 1 ) 1 p 1 V n 0 ε + ε O ˜ ( 1 ) 1 q ,
I ˜ : = n = 1 m = 1 a ˜ m b ˜ n k = 1 s ( U m λ / s + c k V n λ / s ) = n = 1 m = 1 μ m k = 1 s ( U m λ / s + c k V n λ / s ) V n λ ˜ 2 U m 1 λ ˜ 1 υ n V n ε + 1 = n = 1 ϖ s ( λ ˜ 1 , n ) υ n V n ε + 1 k s ( λ ˜ 1 ) n = 1 ( 1 ϑ ( λ ˜ 1 , n ) ) υ n V n ε + 1 = k s ( λ ˜ 1 ) n = 1 υ n V n ε + 1 n = 1 O ( υ n V n ε q + λ 1 + 1 ) = 1 ε k s ( λ ˜ 1 ) 1 V n 0 ε + ε ( O ˜ ( 1 ) O ( 1 ) ) .
If there exists a positive constant K k s ( λ 1 ) , such that (21) is valid when we replace k s ( λ 1 ) by K , then in particular, we have
ε I ˜ < ε K | | a ˜ | | p , Φ λ | | b ˜ | | q , Ψ λ ,
namely
k s ( λ ˜ 1 ) 1 V n 0 ε + ε ( O ˜ ( 1 ) O ( 1 ) ) < K 1 U m 0 ε + ε O ( 1 ) 1 p 1 V n 0 ε + ε O ˜ ( 1 ) 1 q .
It follows that k s ( λ 1 ) K ( ε 0 + ) . Hence, K = k s ( λ 1 ) is the best possible constant factor of (21).
The constant factor k s ( λ 1 ) in (22) is still the best possible. Otherwise, we would reach a contradiction by (25) that the constant factor in (21) is not the best possible.
This completes the proof of the theorem. □
For p > 1 ,
Ψ λ 1 p ( n ) = υ n V n 1 p λ 2 ,
we define the following normed spaces:
l p , Φ λ : = { a = { a m } m = 1 ; | | a | | p , Φ λ < } , l q , Ψ λ : = { b = { b n } n = 1 ; | | b | | q , Ψ λ < } , l p , Ψ λ 1 p : = { c = { c n } n = 1 ; | | c | | p , Ψ λ 1 p < } .
Assuming that a = { a m } m = 1 l p , Φ λ , setting
c = { c n } n = 1 , c n : = m = 1 a m k = 1 s ( U m λ / s + c k V n λ / s ) , n N ,
we can rewrite (22) as:
| | c | | p , Ψ λ 1 p < k s ( λ 1 ) | | a | | p , Φ λ < ,
namely, c l p , Ψ λ 1 p .
Definition 1.
Define a Hilbert-type operator T : l p , Φ λ l p , Ψ λ 1 p as follows: For any a = { a m } m = 1 l p , Φ λ , there exists a unique representation T a = c l p , Ψ λ 1 p . Define the formal inner product of T a and b = { b n } n = 1 l q , Ψ λ as follows:
( T a , b ) : = n = 1 m = 1 a m k = 1 s ( U m λ / s + c k V n λ / s ) b n .
We can express the above results in operator forms as:
( T a , b ) < k s ( λ 1 ) | | a | | p , Φ λ | | b | | q , Ψ λ ,
| | T a | | p , Ψ λ 1 p < k s ( λ 1 ) | | a | | p , Φ λ .
Define the norm of the operator T as follows:
| | T | | : = sup a ( θ ) l p , Φ λ | | T a | | p , Ψ λ 1 p | | a | | p , Φ λ .
Then, by (31), we get that | | T | | k s ( λ 1 ) . Since the constant factor in (31) is the best possible, we have | | T | | = k s ( λ 1 ) .

4. Some Reverses

In the following, we also set
Φ ˜ λ ( m ) = ( 1 θ ( λ 2 , m ) ) U m p ( 1 λ 1 ) 1 μ m p 1 , Ψ ˜ λ ( n ) = ( 1 ϑ ( λ 1 , n ) ) V n q ( 1 λ 2 ) 1 υ n q 1 ( m , n N ) .
For 0 < p < 1 or p < 0 , we still use the formal symbols | | a | | p , Φ λ , | | b | | q , Ψ λ , | | a | | p , Φ ˜ λ and | | b | | q , Ψ ˜ λ .
Theorem 3.
If 0 < p < 1 , m 0 , n 0 N , μ m μ m + 1 ( m { m 0 , m 0 + 1 , } ) , υ n υ n + 1 ( n { n 0 , n 0 + 1 , } ) , U ( ) = V ( ) = , then we have the following equivalent inequalities with the best possible constant factor k s ( λ 1 ) :
I = n = 1 m = 1 a m b n k = 1 s ( U m λ s + c k V n λ s ) > k s ( λ 1 ) | | a | | p , Φ ˜ λ | | b | | q , Ψ λ ,
J = n = 1 υ n V n 1 p λ 2 m = 1 a m k = 1 s ( U m λ s + c k V n λ s ) p 1 p > k s ( λ 1 ) | | a | | p , Φ ˜ λ .
Proof. 
By the reverse Hölder inequality (cf. [44]), we derive the reverses of (2325). Then, by (17), we obtain (33). By (33) and the reverse of (25), we have (32). On the other hand, assuming that (32) is valid, we set b n as in Theorem 1. Then, we get that J p = | | b | | q , Ψ λ q . If J = , then (33) is trivially valid; if J = 0 , then by the reverse of (24) and (17), this is impossible.
Suppose that 0 < J < . By (32), it follows that
| | b | | q , Ψ λ q = J p = I > k s ( λ 1 ) | | a | | p , Φ ˜ λ | | b | | q , Ψ λ ,
| | b | | q , Ψ λ q 1 = J > k s ( λ 1 ) | | a | | p , Φ ˜ λ ,
and then (33) follows, which is equivalent to (32).
For ε ( 0 , p λ 1 ) , we set λ ˜ 1 , λ ˜ 2 , a ˜ m and b ˜ n as in (28). Then by (19), (20) and (16), we find
| | a | | p , Φ ˜ λ | | b | | q , Ψ λ = m = 1 ( 1 θ ( λ 2 , m ) ) μ m U m 1 + ε 1 p n = 1 υ n V n 1 + ε 1 q = m = 1 μ m U m 1 + ε m = 1 O μ m U m 1 + λ 2 + ε q 1 p n = 1 υ n V n 1 + ε 1 q = 1 ε 1 U m 0 ε + ε ( O ( 1 ) O 1 ( 1 ) ) 1 p 1 V n 0 ε + ε O ˜ ( 1 ) 1 q , I ˜ = n = 1 m = 1 a ˜ m b ˜ n k = 1 s ( U m λ / s + c k V n λ / s ) = n = 1 m = 1 μ m k = 1 s ( U m λ / s + c k V n λ / s ) V n λ ˜ 2 U m 1 λ ˜ 1 υ n V n ε + 1 = n = 1 ϖ s ( λ ˜ 1 , n ) υ n V n ε + 1 k s ( λ ˜ 1 ) n = 1 υ n V n ε + 1 = 1 ε k s ( λ ˜ 1 ) 1 V n 0 ε + ε O ˜ ( 1 ) .
If there exists a positive constant K k s ( λ 1 ) , such that (32) is valid when we replace k s ( λ 1 ) by K , then in particular, we have
ε I ˜ > ε K | | a ˜ | | p , Φ ˜ λ | | b ˜ | | q , Ψ λ ,
namely,
k s ( λ ˜ 1 ) 1 V n 0 ε + ε O ˜ ( 1 ) > K 1 U m 0 ε + ε ( O ( 1 ) O 1 ( 1 ) ) 1 p 1 V n 0 ε + ε O ˜ ( 1 ) 1 q .
It follows that k s ( λ 1 ) K ( ε 0 + ) . Hence, K = k s ( λ 1 ) is the best possible constant factor of (32). The constant factor k s ( λ 1 ) in (33) is still the best possible. Otherwise, we would reach a contradiction by the reverse of (25) that the constant factor in (32) is not the best possible.
This completes the proof of the theorem. □
Theorem 4.
If p < 0 , m 0 , n 0 N , μ m μ m + 1 ( m { m 0 , m 0 + 1 , } ) , υ n υ n + 1 ( n { n 0 , n 0 + 1 , } ) , U ( ) = V ( ) = , then we have the following equivalent inequalities with the best possible constant factor k s ( λ 1 ) :
I = n = 1 m = 1 a m b n k = 1 s ( U m λ s + c k V n λ s ) > k s ( λ 1 ) | | a | | p , Φ λ | | b | | q , Ψ ˜ λ , J 1 : = n = 1 V n p λ 2 1 υ n ( 1 ϑ ( λ 1 , n ) ) p 1 m = 1 a m k = 1 s ( U m λ s + c k V n λ s ) p 1 p
> k s ( λ 1 ) | | a | | p , Φ λ .
Proof. 
By the reverse Hölder inequality with weight (cf. [44]), since p < 0 , by (18), we have
m = 1 a m k = 1 s ( U m λ / s + c k V n λ / s ) p = m = 1 1 k = 1 s ( U m λ s + c k V n λ s ) U m ( 1 λ 1 ) / q υ n 1 / p a m V n ( 1 λ 2 ) / p μ m 1 / q V n ( 1 λ 2 ) / p μ m 1 / q U m ( 1 λ 1 ) / q υ n 1 / p p m = 1 1 k = 1 s ( U m λ / s + c k V n λ / s ) U m ( 1 λ 1 ) p / q υ n V n 1 λ 2 μ m p / q a m p × m = 1 1 k = 1 s ( U m λ / s + c k V n λ / s ) V n ( 1 λ 2 ) ( q 1 ) μ m U m 1 λ 1 υ n q 1 p 1 = ( ϖ s ( λ 1 , n ) ) p 1 V n p λ 2 1 υ n m = 1 1 k = 1 s ( U m λ / s + c k V n λ / s ) U m ( 1 λ 1 ) ( p 1 ) υ n V n 1 λ 2 μ m p 1 a m p ( k s ( λ 1 ) ) p 1 ( 1 ϑ ( λ 1 , n ) ) p 1 V n p λ 2 1 υ n × m = 1 1 k = 1 s ( U m λ / s + c k V n λ / s ) U m ( 1 λ 1 ) ( p 1 ) υ n V n 1 λ 2 μ m p 1 a m p , J 1 ( k s ( λ 1 ) ) 1 q n = 1 m = 1 1 k = 1 s ( U m λ / s + c k V n λ / s ) U m ( 1 λ 1 ) ( p 1 ) υ n V n 1 λ 2 μ m p 1 a m p 1 p = ( k s ( λ 1 ) ) 1 q m = 1 n = 1 1 k = 1 s ( U m λ / s + c k V n λ / s ) U m ( 1 λ 1 ) ( p 1 ) υ n V n 1 λ 2 μ m p 1 a m p 1 p = ( k s ( λ 1 ) ) 1 q m = 1 ω s ( λ 2 , m ) U m p ( 1 λ 1 ) 1 μ m p 1 a m p 1 p .
Then by (15), we obtain (37).
By the reverse Hölder inequality (cf. [44]), we have
I = n = 1 V n λ 2 1 p υ n 1 / p ( 1 ϑ ( λ 1 , n ) ) 1 / q m = 1 a m k = 1 s ( U m λ / s + c k V n λ / s )
× ( 1 ϑ ( λ 1 , n ) ) 1 q V n 1 p λ 2 υ n 1 / p b n J 1 | | b | | q , Ψ ˜ λ .
Then, by (37), we deduce (36). On the other hand, assuming that (36) is valid, we set b n as follows:
b n : = V n p λ 2 1 υ n ( 1 ϑ ( λ 1 , n ) ) p 1 m = 1 a m k = 1 s ( U m λ / s + c k V n λ / s ) p 1 , n N .
Then, we obtain that J 1 p = | | b | | q , Ψ ˜ λ q . If J 1 = , then (37) is trivially valid; if J 1 = 0 , then by (15) and (38), this is impossible. Suppose that 0 < J 1 < . By (36), it follows that
| | b | | q , Ψ ˜ λ q = J 1 p = I > k s ( λ 1 ) | | a | | p , Φ λ | | b | | q , Ψ ˜ λ , | | b | | q , Ψ ˜ λ q 1 = J 1 > k s ( λ 1 ) | | a | | p , Φ λ ,
and then (37) follows, which is equivalent to (36).
For ε ( 0 , | p | λ 1 ) , we set λ ˜ 1 = λ 1 + ε q ( > 0 ) , λ ˜ 2 = λ 2 ε q ( ( 0 , 1 ) ) , and
a ˜ m : = U m λ ˜ 1 1 ε μ m = U m λ 1 ε p 1 μ m , b ˜ n = V n λ ˜ 2 1 υ n = V n λ 2 ε q 1 υ n .
Then, by (17), (19) and (20), we have
| | a ˜ | | p , Φ λ | | b ˜ | | q , Ψ ˜ λ = m = 1 μ m U m 1 + ε 1 p n = 1 ( 1 ϑ ( λ 1 , n ) ) υ n V n 1 + ε 1 q = m = 1 μ m U m 1 + ε 1 p n = 1 υ n V n 1 + ε n = 1 O υ n V n 1 + λ 1 + ε 1 q = 1 ε 1 U m 0 ε + ε O ( 1 ) 1 p 1 V n 0 ε + ε ( O ˜ ( 1 ) O 1 ( 1 ) ) 1 q , I ˜ = n = 1 m = 1 a ˜ m b ˜ n k = 1 s ( U m λ / s + c k V n λ / s ) = m = 1 n = 1 υ n k = 1 s ( U m λ / s + c k V n λ / s ) U m λ ˜ 1 V n 1 λ ˜ 2 μ m U m 1 + ε = m = 1 ω s ( λ ˜ 2 , m ) μ m U m 1 + ε k s ( λ ˜ 1 ) n = 1 μ m U m 1 + ε = 1 ε k s ( λ ˜ 1 ) 1 U m 0 ε + ε O ( 1 ) .
If there exists a positive constant K k s ( λ 1 ) , such that (36) is valid when we replace k s ( λ 1 ) by K , then in particular, we have
ε I ˜ > ε K | | a ˜ | | p , Φ λ | | b ˜ | | q , Ψ ˜ λ ,
namely,
k s ( λ ˜ 1 ) 1 U m 0 ε + ε O ( 1 ) > K 1 U m 0 ε + ε O ( 1 ) 1 p 1 V n 0 ε + ε ( O ˜ ( 1 ) O 1 ( 1 ) ) 1 q .
It follows that k s ( λ 1 ) K ( ε 0 + ) . Hence, K = k s ( λ 1 ) is the best possible constant factor of (36).
The constant factor k s ( λ 1 ) in (37) is still the best possible. Otherwise, we would reach a contradiction by (39) that the constant factor in (36) is not the best possible.
This completes the proof of the theorem. □
Remark 1.(i) For μ i = υ j = 1 ( i , j N ) , (21) reduces to (7).
(ii) For
s = λ = c 1 = 1 , λ 1 = 1 q , λ 2 = 1 p ,
(21) reduces to (4); for s = λ = c 1 = 1 , λ 1 = 1 p , λ 2 = 1 q , (21) reduces to the dual form of (4) as follows:
m = 1 n = 1 a m b n U m + V n < π sin ( π / p ) m = 1 U m p 2 μ m p 1 a m p 1 p n = 1 V n q 2 υ n q 1 b n q 1 q .
(iii) For p = q = 2 , both (4) and (40) reduce to
m = 1 n = 1 a m b n U m + V n < π m = 1 a m 2 μ m n = 1 b n 2 υ n 1 2 .

5. Conclusions

In the present paper, making use of weight coefficients as well as real/complex analytic methods, a Hardy–Hilbert-type inequality with a best possible constant factor and multiparameters and the equivalent forms are established in Theorems 1 and 2. Reverses, operator expression with the norm, and a few particular cases are also considered in Theorems 3 and 4, Definition 1, and Remark 1. The lemmas and theorems provide an extensive account of this type of inequality.

Author Contributions

Writing—original draft preparation, B.Y., M.T.R. and A.R.; project administration, B.Y., M.T.R. and A.R. All authors contributed equally in all stages of preparation of this work. All authors have read and agreed to the published version of the manuscript.

Funding

B. Yang: This work is supported by the National Natural Science Foundation of China (No. 61772140) and the Characteristic Innovation Project of Guangdong Provincial Colleges and Universities in 2020 (No. 2020KTSCX088). We are grateful for this help. A. Raigorodskii: This author acknowledges the Russian Federation Government for the financial support of his study: his research on these results was carried out with the support of megagrant number 075-15-2019-1926. His research on these results was also supported in the framework of the grant “Leading scientific schools” number NSh-2540.2020.1 (075-15-2020-417).

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Data Availability Statement

Not applicable.

Conflicts of Interest

The authors declare that they do not have any conflict of interest.

References

  1. Hardy, G.H.; Littlewood, J.E.; Polya, G. Inequalities; Cambridge University Press: Cambridge, UK, 1934. [Google Scholar]
  2. Mitrinovic, D.S.; Pecaric, J.E.; Fink, A.M. Inequalities Involving Functions and Their Integrals and Derivatives; Kluwer Acaremic Publishers: Boston, FL, USA, 1991. [Google Scholar]
  3. Mitrinovic, D.S.; Pecaric, J.E.; Fink, A.M. Classical and New Inequalities in Analysis; Springer: Berlin/Heidelberg, Germany, 1993. [Google Scholar]
  4. Mitrinović, D.S. Analytic Inequalities; Springer: Berlin/Heidelberg, Germany, 1970. [Google Scholar]
  5. Milovanović, G.V. (Ed.) Recent Progress in Inequalities; Springer: Berlin/Heidelberg, Germany, 1998. [Google Scholar]
  6. Yang, B.C. The Norm of Operator and Hilbert-Type Inequalities; Science Press: Beijing, China, 2009. [Google Scholar]
  7. Burtseva, E.; Lundberg, S.; Persson, L.-E.; Samko, N. Multi-dimensional Hardy type inequalities in Holder spaces. J. Math. Inequalities 2018, 12, 719–729. [Google Scholar] [CrossRef][Green Version]
  8. Jaksetic, J.; Pecaric, J.; Kalamir, K.S. Further improvement of an extension of Holder-type inequality. Math. Inequalities Appl. 2019, 22, 1161–1175. [Google Scholar] [CrossRef][Green Version]
  9. Batbold, T.; Laith, E.A. A new form of Hilbert integral inequality. J. Math. Inequalities 2018, 12, 379–390. [Google Scholar] [CrossRef][Green Version]
  10. Fabelurin, O.O.; Oguntuase, J.A.; Persson, L.E. Multidimensional Hardy-type inequality on time scales with variable exponents. J. Math. Inequalities 2019, 13, 725–736. [Google Scholar] [CrossRef][Green Version]
  11. Basci, Y. Baleanu, D. Hardy-type inequalities within fractional derivatives without singular kernel. J. Inequalities Appl. 2018, 2018, 304. [Google Scholar] [CrossRef]
  12. Zhao, C.J.; Cheung, W.S. Reverse Hilbert type inequalities. J. Math. Inequalities 2019, 13, 855–866. [Google Scholar] [CrossRef]
  13. Yang, B.C. On Hilbert’s integral inequality. J. Math. Anal. Appl. 1998, 220, 778–785. [Google Scholar]
  14. Yang, B.C. On a more accurate multidimensional Hilbert-type inequality with parameters. Math. Inequalities Appl. 2015, 18, 429–441. [Google Scholar] [CrossRef]
  15. Yang, B.C.; Brnetic, I.; Krnic, M.; Pecaric, J.E. Generalization of Hilbert and Hardy-Hilbert integral inequalities. Math. Ineq. Appl. 2005, 8, 259–272. [Google Scholar]
  16. Krnic, M.; Pecaric, J.E. Hilbert’s inequalities and their reverses. Publ. Math. Debrecen 2005, 67, 315–331. [Google Scholar]
  17. Yang, B.C.; Rassias, T.M. On the way of weight coefficient and research for Hilbert-type inequalities. Math. Inequalities Appl. 2003, 6, 625–658. [Google Scholar] [CrossRef][Green Version]
  18. Azar, L. On some extensions of Hardy-Hilbert’s inequality and Applications. J. Inequalities Appl. 2009, 2009, 546829. [Google Scholar] [CrossRef]
  19. Arpad, B.; Choonghong, O. Best constant for certain multilinear integral operator. J. Inequalities Appl. 2006, 2006, 28582. [Google Scholar]
  20. Kuang, J.C.; Debnath, L. On Hilbert’s type inequalities on the weighted Orlicz spaces. Pacific J. Appl. Math. 2007, 1, 95–103. [Google Scholar]
  21. Zhong, W.Y. The Hilbert-type integral inequality with a homogeneous kernel of Lambda-degree. J. Inequalities Appl. 2008, 2008, 917392. [Google Scholar] [CrossRef][Green Version]
  22. Hong, Y. On Hardy–Hilbert integral inequalities with some parameters. J. Ineq. Pure Appl. Math. 2005, 6, 92. [Google Scholar]
  23. Krnic, M.; Pecaric, J.; Vukovic, P. On some higher-dimensional Hilbert’s and Hardy-Hilbert’s type integral inequalities with parameters. Math. Inequalities Appl. 2008, 11, 701–716. [Google Scholar] [CrossRef][Green Version]
  24. Krnic, M.; Vukovic, P. On a multidimensional version of the Hilbert-type inequality. Anal. Math. 2012, 38, 291–303. [Google Scholar] [CrossRef]
  25. Baleanu, D.; Krnic, M.; Vukovic, P. A class of fractal Hilbert-type inequalities obtained via Cantor-type spherical coordinates. Math. Methods Appl. Sci. 2021, 44, 6195–6208. [Google Scholar] [CrossRef]
  26. Krnić, M.; Pecaric, J.E. Extension of Hilbert’s inequality. J. Math. Anal. Appl. 2006, 324, 150–160. [Google Scholar] [CrossRef][Green Version]
  27. Krnic, M. On the multidimensional Hilbert-type inequalities involving the hardy operator. Filomat 2012, 26, 845–857. [Google Scholar] [CrossRef][Green Version]
  28. Krnić, M.; Vuković, P. Multidimensional Hilbert-type inequalities obtained via local fractional calculus. Acta Appl. Math. 2020, 169, 667–680. [Google Scholar] [CrossRef]
  29. Adiyasuren, V.; Batbold, T.; Krnić, M. Half-discrete Hilbert-type inequalities with mean operators, the best constants, and applications. Appl. Math. Comput. 2014, 231, 148–159. [Google Scholar] [CrossRef]
  30. Adiyasuren, V.; Batbold, T.; Krnić, M. On several new Hilbert-type inequalities involving means operators. Acta Math. Sin. Engl. Ser. 2013, 29, 1493–1514. [Google Scholar] [CrossRef]
  31. Brnetić, I.; Krnić, M.; Pecarić, J. Multiple Hilbert and Hardy–Hilbert inequalities with non-conjugate parameters. Bull. Aust. Math. Soc. 2005, 71, 447–457. [Google Scholar] [CrossRef][Green Version]
  32. Peric, I.; Vukovic, P. Multiple Hilbert’s type inequalities with a homogeneous kernel. Banach J. Math. 2011, 5, 33–43. [Google Scholar] [CrossRef]
  33. Adiyasuren, V.; Batbold, T.; Mario, K.M. Multiple Hilbert-type inequalities involving some differential operators. Banach J. Math. Anal. 2016, 10, 320–337. [Google Scholar] [CrossRef]
  34. Huang, Q.L. A new extension of Hardy–Hilbert-type inequality. J. Inequalities Appl. 2015, 2015, 397. [Google Scholar] [CrossRef][Green Version]
  35. He, B. A multiple Hilbert-type discrete inequality with a new kernel and best possible constant factor. J. Math. Anal. Appl. 2015, 431, 889–902. [Google Scholar] [CrossRef]
  36. Li, Y.J.; He, B. On inequalities of Hilbert’s type. Bull. Aust. Math. Soc. 2007, 76, 1–13. [Google Scholar] [CrossRef][Green Version]
  37. Rassias, M.T.; Yang, B. On a Hilbert-type integral inequality related to the extended Hurwitz zeta function in the whole plane. Acta Appl. Math. 2019, 160, 67–80. [Google Scholar] [CrossRef]
  38. Rassias, M.T.; Yang, B. On a Hilbert-type integral inequality in the whole plane related to the extended Riemann zeta function. Complex Anal. Oper. Theory 2019, 13, 1765–1782. [Google Scholar] [CrossRef]
  39. Rassias, M.T.; Yang, B.; Raigorodskii, A. Two kinds of the reverse Hardy-type integral inequalities with the equivalent forms related to the extended Riemann zeta function. Appl. Anal. Discrete Math. 2018, 12, 273–296. [Google Scholar] [CrossRef]
  40. Rassias, M.T.; Yang, B.C. On an equivalent property of a reverse Hilbert-type integral inequality related to the extended Hurwitz-zeta function. J. Math. Inequalities 2019, 13, 315–334. [Google Scholar] [CrossRef][Green Version]
  41. Rassias, M.T.; Yang, B.C.; Raigorodskii, A. On the reverse Hardy-type integral inequalities in the whole plane with the extended Riemann-zeta function. J. Math. Inequalities 2020, 14, 525–546. [Google Scholar] [CrossRef]
  42. Rassias, M.T.; Yang, B. Equivalent properties of a Hilbert-type integral inequality with the best constant factor related to the Hurwitz zeta function. Ann. Funct. Anal. 2018, 9, 282–295. [Google Scholar] [CrossRef]
  43. Pan, Y.L.; Wang, H.T.; Wang, F.T. On Complex Functions; Science Press: Beijing, China, 2006. [Google Scholar]
  44. Kuang, J.C. Applied Inequalities; Shangdong Science Technic Press: Jinan, China, 2004. [Google Scholar]
Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Share and Cite

MDPI and ACS Style

Yang, B.; Rassias, M.T.; Raigorodskii, A. On an Extension of a Hardy–Hilbert-Type Inequality with Multi-Parameters. Mathematics 2021, 9, 2432. https://doi.org/10.3390/math9192432

AMA Style

Yang B, Rassias MT, Raigorodskii A. On an Extension of a Hardy–Hilbert-Type Inequality with Multi-Parameters. Mathematics. 2021; 9(19):2432. https://doi.org/10.3390/math9192432

Chicago/Turabian Style

Yang, Bicheng, Michael Th. Rassias, and Andrei Raigorodskii. 2021. "On an Extension of a Hardy–Hilbert-Type Inequality with Multi-Parameters" Mathematics 9, no. 19: 2432. https://doi.org/10.3390/math9192432

Note that from the first issue of 2016, MDPI journals use article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop