Open Access

A note on l p norms of weighted mean matrices

Journal of Inequalities and Applications20122012:110

https://doi.org/10.1186/1029-242X-2012-110

Received: 11 January 2012

Accepted: 19 May 2012

Published: 19 May 2012

Abstract

We present some results concerning the l p norms of weighted mean matrices. These results can be regarded as analogues to a result of Bennett concerning weighted Carleman's inequalities.

2000 Mathematics Subject Classification: Primary 47A30.

Keywords

Carleman's inequality Hardy's inequality weighted mean matrices

1. Introduction

Let throughout that p 0 , 1 p + 1 q = 1 . For p ≥ 1, let l p be the Banach space of all complex sequences a = (a n )n ≥ 1with norm
a p : = n = 1 a n p 1 / p < .
The celebrated Hardy's inequality [1], Theorem 326] asserts that for p > 1,
n = 1 1 n k = 1 n a k p p p - 1 p n = 1 a n p .
(1.1)
Hardy's inequality can be regarded as a special case of the following inequality:
C a p p = n = 1 k = 1 c n , k a k p U n = 1 a n p ,
(1.2)
in which C = (cn,k) and the parameter p > 1 are assumed fixed, and the estimate is to hold for all complex sequences a l p . The l p operator norm of C is then defined as
C p , p = sup a p = 1 C a p .

It follows that inequality (1.2) holds for any a l p when U1/p≥ ||C||p,pand fails to hold for some a l p when U1/p< ||C||p,p. Hardy's inequality thus asserts that the Cesáro matrix operator C, given by cn,k= 1/n, kn and 0 otherwise, is bounded on l p and has norm ≤ p/(p - 1). (The norm is in fact p/(p - 1).)

We say a matrix A = (an,k) is a lower triangular matrix if an,k= 0 for n < k and a lower triangular matrix A is a summability matrix if an,k≥ 0 and k = 1 n a n , k = 1 . We say a summability matrix A is a weighted mean matrix if its entries satisfy:
a n , k = λ k / Λ n , 1 k n ; Λ n = i = 1 n λ i , λ i 0 , λ 1 > 0 .
(1.3)

Hardy's inequality (1.1) now motivates one to determine the l p operator norm of an arbitrary summability or weighted mean matrix A. Gao [2] proved the following result:

Theorem 1.1. Let 1 < p < ∞ be fixed. Let A be a weighted mean matrix given by (1.3). If for any integer n ≥ 1, there exists a positive constant 0 < L < p such that
Λ n + 1 λ n + 1 Λ n λ n 1 - L λ n p Λ n 1 - p + L p ,
(1.4)

then ||A||p,pp/(p-L).

It is easy to see that the above result implies the following well-known result of Cartlidge [3] (see also [4], p. 416, Theorem C]):

Theorem 1.2. Let 1 < p < ∞ be fixed. Let A be a weighted mean matrix given by (1.3). If
L = sup n Λ n + 1 λ n + 1 - Λ n λ n < p ,
(1.5)

then ||A||p,pp/(p-L).

The above result of Cartlidge are very handy to use when determining l p norms of certain weighted mean matrices. We refer the readers to the articles [2, 57] for more recent developments in this area.

We note here that by a change of variables a k a k 1 / p in (1.1) and on letting p → +∞, one obtains the following well-known Carleman's inequality [8], which asserts that for convergent infinite series ∑a n with non-negative terms, one has
n = 1 ( k = 1 n a k ) 1 n e n = 1 a n ,

with the constant e being best possible.

It is then natural to study the following weighted version of Carleman's inequality:
n = 1 k = 1 n a k λ k / Λ n E n = 1 a n ,
(1.6)
where the notations are as in (1.3). The task here is to determine the best constant E so that inequality (1.6) holds for any convergent infinite series ∑a n with non-negative terms. Note that Cartlidge's result (Theorem 1.2) implies that when (1.5) is satisfied, then for any a l p , one has
n = 1 k = 1 n λ k a k Λ n p p p - L p n = 1 a n p .
(1.7)

Similar to our discussions above, by a change of variables a k a k 1 / p in (1.7) and on letting p → +∞, one obtains inequality (1.6) with E = e L as long as (1.5) is satisfied with p replaced by +∞ there.

Note that (1.5) can be regarded as the case p → 1+ of (1.4) while the case p → +∞ of (1.4) suggests the following result:

Corollary 1.1. Suppose that
M = sup n Λ n λ n log Λ n + 1 / λ n + 1 Λ n / λ n < + ,

then inequality (1.6) holds with E = e M .

In fact, the above corollary is a consequence of the following nice result of Bennett (see the proof of [5, Theorem 13]):

Theorem 1.3. Inequality (1.6) holds with
E = sup n Λ n + 1 λ n + 1 k = 1 n λ k Λ k λ k / Λ n .

It is shown in the proof of Theorem 13 in [5] that Corollary 1.1 follows from the above theorem. It is also easy to see that ML for L defined by (1.5) so that Corollary 1.1 provides a better result than what one can infer from Cartlidge's result as discussed above.

Note that the bound given in Theorem 1.3 is global in the sense that it involves all the λ n 's and it implies the local version Corollary 1.1, in which only the terms Λ n n and Λn+1/λn+1are involved. It is then natural to ask whether one can obtain a similar result for the l p norms for p > 1 so that it implies the local version Theorem 1.1. It is our goal in this note to present one such result and as our result is motivated by the result of Bennett, we will first study the limiting case p → +∞, namely weighted Carleman's inequalities in the next section before we move on to the l p cases in Section 3.

2. A discussion on weighted Carleman's inequalities

In this section, we study the weighted Carleman's inequalities. Our goal is to give a different proof of Theorem 1.3 than that given in [5] and discuss some variations of it. It suffices to consider the cases of (1.6) with the infinite summations replaced by any finite summations, say from 1 to N ≥ 1 here. Our starting point is the following result of Pečarić and Stolarsky [[9], (2.4)], which is an outgrowth of Redheffer's approach in [10]:
n = 1 N Λ n ( b n - 1 ) G n + Λ N G N n = 1 N λ n a n b n Λ n / λ n ,
(2.1)
where b is any positive sequence and
G n = k = 1 n a k λ k / Λ n .
We now discard the last term on the left-hand side of (2.1) and make a change of variables λ n a n b n Λ n / λ n a n to recast inequality (2.1) as
n = 1 N Λ n ( b n - 1 ) k = 1 n λ k - λ k / Λ n k = 1 n b k - Λ k / Λ n G n n = 1 N a n .
Now, a further change of variables b n λn+1b n n allows us to recast the above inequality as
n = 1 N Λ n b n λ n - 1 λ n + 1 k = 1 n b k - Λ k / Λ n G n n = 1 N a n .
(2.2)
If one now chooses b n = (Λn+1n+1)/(Λ n n ) such that Λ n (b n n - 1/λn+1) = 1, then one gets immediately the following:
n = 1 N λ n + 1 Λ n + 1 k = 1 n Λ k λ k λ k / Λ n G n n = 1 N a n ,

from which the assertion of Theorem 1.3 can be readily deduced.

Another natural choice for the values of b n 's is to set k = 1 n b k - Λ k / Λ n = e - M . From this we see that b n = e M λ n / Λ n and substituting these values for b n 's we obtain via (2.2):
n = 1 N Λ n e M λ n / Λ n λ n - 1 λ n + 1 G n e M n = 1 N a n .
(2.3)

One checks easily that the above inequality implies Corollary 1.1.

We now consider a third choice for the b n 's by setting b n = e ( Λ n + 1 / λ n + 1 - Λ n / λ n ) / ( Λ n / λ n ) and it follows from this and (2.2) that
n = 1 N Λ n e ( Λ n + 1 / λ n + 1 - Λ n / λ n ) / ( Λ n / λ n ) λ n - 1 λ n + 1 e - k = 1 n λ k Λ n ( Λ k + 1 λ k + 1 - Λ k λ k ) G n n = 1 N a n ,

from which we deduce the following

Corollary 2.1. Suppose that
M = sup n k = 1 n λ k Λ n Λ k + 1 λ k + 1 - Λ k λ k < + ,

then inequality (1.6) holds with E = e M .

We point out here that the above corollary also provides a better result than what one can infer from Cartlidge's result and it also follows from Theorem 1.3.

We note here that the optimal choice for the b n 's will be to choose them to satisfy
Λ n b n λ n - 1 λ n + 1 k = 1 n b k - Λ k / Λ n = e - L .
In general, it is difficult to solve for the b n 's from the above equations. But we can solve b1 to get b1 = e L λ1/((e L - 1)λ2) and if we set b n = (Λn+1n+1)/(Λ n n ) for n ≥ 2, we can then deduce from (2.2) that
e - L G 1 + n = 2 N Λ 2 ( e L - 1 ) λ 1 e L λ 1 / Λ N λ n + 1 Λ n + 1 k = 1 n Λ k λ k λ k / Λ n G n n = 1 N a n .
Note that this gives an improvement upon Theorem 1.3 as long as λ22 > e-L. Similarly, one obtains
G 1 + n = 2 N λ 2 ( e L - 1 ) λ 1 λ 1 / Λ N Λ n e M λ n / Λ n λ n - 1 λ n + 1 G n e M n = 1 N a n .

3. The l p cases

We now return to the discussions on the general l p cases. Again it suffices to consider the cases of (1.7) with the infinite summations replaced by any finite summations, say from 1 to N ≥ 1 here. We may also assume that a n ≥ 0 for all n. With the discussions of the previous section in mind, here we seek for an l p version of (2.2). Fortunately, this is available by noting that it follows from inequality (4.3) of [7] that
n = 1 N k = 1 n w k - ( p - 1 ) w n p - 1 λ n p - w n + 1 p - 1 λ n + 1 p Λ n p A n p n = 1 N a n p ,
(3.1)
where w n 's are positive parameters and
A n = k = 1 n λ k a k Λ n .
By a change of variables w n λ n w n 1 / ( p - 1 ) , we can recast inequality (3.1) as
n = 1 N k = 1 n λ k w k 1 / ( p - 1 ) Λ n - ( p - 1 ) w n λ n - w n + 1 λ n + 1 Λ n A n p n = 1 N a n p ,
With another change of variables, w n /wn+1 b n , we can further recast the above inequality as
n = 1 N k = 1 n λ k i = k n b i 1 / ( p - 1 ) Λ n - ( p - 1 ) b n λ n - 1 λ n + 1 Λ n A n p n = 1 N a n p .
(3.2)

Note that if one makes a change of variables a n p a n , then inequality (2.2) follows from the above inequality upon letting p → +∞.

One can then deduce Theorem 1.1 by choosing b n = (1-L λ n /(p Λ n ))-(p-1)in (3.2) (see [2]). It is easy to see that this gives back inequality (2.3) upon letting p → +∞ by a change of variables a n p a n and setting L = M. We note here that the b n 's are so chosen so that the following relations are satisfied:
k = 1 n λ k i = k n b i 1 / ( p - 1 ) Λ n = p p - L .
(3.3)
Now, to get the l p analogues of Theorem 1.3, we just need to note that in the p → +∞ case, one obtains Corollary 1.1 by setting k = 1 n b k - Λ k / Λ n = e - M and the conclusion of Corollary 1.1 follows by requiring that Λ n (b n n - 1/λn+1) ≥ 1 for the so chosen b n 's. If one instead chooses the b n 's so that the conditions Λ n (b n n - 1/λn+1) = 1 are satisfied, then Theorem 1.3 will follow. Now, in the l p cases, the choice of the b n 's so that the conditions (3.3) are satisfied implies Theorem 1.1 as (1.4) implies that for the so chosen b n 's,
b n λ n - 1 λ n + 1 Λ n 1 - L p .
Thus, in order to obtain an result analogue to Theorem 1.3 for the l p cases, we are then motivated to take the b n 's so that the following conditions are satisfied:
b n λ n - 1 λ n + 1 Λ n = 1 - L p .
(3.4)

We then easily deduce the following l p analogue of Theorem 1.3:

Theorem 3.1. Let 1 < p < ∞ be fixed. Let A be a weighted mean matrix given by (1.3). If for any integer n ≥ 1, there exists a positive constant 0 < L < p such that
k = 1 n λ k Λ n i = k n Λ i + 1 / λ i + 1 - L / p Λ i / λ i 1 / ( p - 1 ) p p - L ,

then ||A||p,pp/(p-L).

It is easy to see by induction that Theorem 3.1 implies Theorem 1.1. Of course one should really choose the b n 's so that the following relations are satisfied:
k = 1 n λ k i = k n b i 1 / ( p - 1 ) Λ n - ( p - 1 ) b n λ n - 1 λ n + 1 Λ n = p p - L - p .

In general it is difficult to determine the b n 's this way but one can certainly solve for b1 and by choosing other b n 's so that (3.4) are satisfied, one can obtain a slightly better result than Theorem 3.1, we shall leave the details to the reader.

We note here that the choice of the b n 's satisfying (3.4) corresponds to the following choice for the a n 's in Section 4 of [2] (these a n 's are not to be confused with the a n 's used in the rest of the article):
a n = Λ n + 1 / λ n + 1 - L / p Λ n / λ n 1 / ( p - 1 ) a n + 1 , a 1 = 1 .
We also note here that in the case of λ n = L = 1, on choosing b n 's to satisfy (3.4), we obtain via (3.2) that
n = 1 k = 1 n i = k n ( 1 + ( 1 - 1 / p ) / i ) 1 / ( p - 1 ) n - ( p - 1 ) 1 n k = 1 n a k p p p - 1 n = 1 a n p .

This gives an improvement of Hardy's inequality (1.1).

Declarations

Authors’ Affiliations

(1)
Division of Mathematical Sciences, School of Physical and Mathematical Sciences, Nanyang Technological University

References

  1. Hardy GH, Littlewood JE, Pólya G: Inequalities. Cambridge University Press, Cambridge; 1952.Google Scholar
  2. Gao P: On l p norms of weighted mean matrices. Math Z 2010, 264: 829–848. 10.1007/s00209-009-0490-2MathSciNetView ArticleGoogle Scholar
  3. Cartlidge JM: Weighted mean matrices as operators on l p . Indiana University; 1978.Google Scholar
  4. Bennett G: Some elementary inequalities. Q J Math Oxford Ser 1987, 38(2):401–425.View ArticleGoogle Scholar
  5. Bennett G: Sums of powers and the meaning of l p . Houston J Math 2006, 32: 801–831.MathSciNetGoogle Scholar
  6. Gao P: A note on Hardy-type inequalities. Proc Am Math Soc 2005, 133: 1977–1984. 10.1090/S0002-9939-05-07964-5View ArticleGoogle Scholar
  7. Gao P: Hardy-type inequalities via auxiliary sequences. J Math Anal Appl 2008, 343: 48–57. 10.1016/j.jmaa.2008.01.024MathSciNetView ArticleGoogle Scholar
  8. Carleman T: Sur les fonctions quasi-analytiques. In Proc 5th Scand Math Congress. Helsingfors, Finland; 1923:181–196.Google Scholar
  9. Pečarić J, Stolarsky K: Carleman's inequality: history and new generalizations. Aequationes Math 2001, 61: 49–62. 10.1007/s000100050160MathSciNetView ArticleGoogle Scholar
  10. Redheffer RM: Recurrent inequalities. Proc Lond Math Soc 1967, 17(3):683–699.MathSciNetView ArticleGoogle Scholar

Copyright

© Gao; licensee Springer. 2012

This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.