Journal of Inequalities in Pure and Applied Mathematics http://jipam.vu.edu.au/ Volume 3, Issue 5, Article 75, 2002 A NEW INEQUALITY SIMILAR TO HILBERT’S INEQUALITY BICHENG YANG D EPARTMENT OF M ATHEMATICS G UANGDONG E DUCATION C OLLEGE , G UANGZHOU , G UANGDONG 510303, P EOPLE ’ S R EPUBLIC OF C HINA . bcyang@pub.guangzhou.gd.cn Received 17 May, 2001; accepted 17 June, 2002 Communicated by J.E. Pečarić A BSTRACT. In this paper, we build a new inequality similar to Hilbert’s inequality with a best constant factor. As an application, we consider its equivalent form. Key words and phrases: Hilbert’s inequality, Weight coefficient, Cauchy’s inequality. 2000 Mathematics Subject Classification. 26D15. 1. I NTRODUCTION P 2 If 0 < n=0 a2n < ∞ and 0 < ∞ n=0 bn < ∞, then the famous Hilbert’s inequality (see Hardy et al. [1]) is given by ! 12 ∞ X ∞ ∞ ∞ X X X am b n (1.1) <π a2n b2n , m + n + 1 n=0 m=0 n=0 n=0 P∞ where the constant factor π is the best possible. Recently, Yang and Debnath [2, 3] and Yang [4, 5] gave (1.1) some extensions and improvements, and Kuang and Debnath [6] considered its strengthened versions and generalizations. The major objective of this paper is to build a new inequality similar to (1.1), which relates to the double series form as ∞ X ∞ ∞ X ∞ X X am b n am b n (1.2) = . ln m + ln n + 1 n=1 m=1 ln emn n=1 m=1 For this, we must estimate the following weight coefficient √ 12 ∞ X 1 ln en √ (n ∈ N ), (1.3) ω(n) = m ln emn ln em m=1 and do some preparatory works. ISSN (electronic): 1443-5756 c 2002 Victoria University. All rights reserved. 044-01 2 B ICHENG YANG 2. S OME L EMMAS Let f have its first four derivatives on [1, ∞) and (−1)n f (n) (x) > 0 (n = 0, . . . , 4), and f (x), f 0 (x) −→ 0 (x → ∞), then (see [6, (2.1)]) Z ∞ ∞ X 1 1 f (x)dx + f (1) − f 0 (1). (2.1) f (k) < 2 12 1 k=1 Lemma 2.1. For n ∈ N, define R(n) as Z 1√ 2 ln en 1 1 2 1 − . (2.2) R(n) = √ 1 du − 1 3 ln en 12(ln en)2 (1 + u)u 2 (2 ln en) 2 0 Then we have R(n) > 0(n ∈ N ). Proof. Integrating by parts, we have Z 1√ Z 1√ 2 ln en 2 ln en 1 1 1 du 2 1 du = 2 (1 + u) (1 + u)u 2 0 0 Z 1√ 2 ln en √ 1 1 1 1 +2 u2 du = (2 ln en) 2 ln en (1 + u)2 0 Z 1 √ 1 1 4 2 ln √en 1 = (2 ln en) 2 + du3/2 (1 + u)2 ln en 3 0 √ √ 1 1 1 1 1 = (2 ln en) 2 + (2 ln en) 2 ln en 3 (ln en)2 Z 1 8 2 ln √en 3/2 1 u du + (1 + u)3 3 0 √ √ 1 1 1 1 1 + (2 ln en) 2 . > (2 ln en) 2 ln en 3 (ln en)2 Hence by (2.2), we have R(n) > 1 1 2 1 1 1 + − − = + > 0. 2 2 ln en 3(ln en) 3 ln en 12(ln en) 3 ln en 4(ln en)2 The lemma is thus proved. Lemma 2.2. If ω(n) is defined by (1.3), then ω(n) < π, for n ∈ N . Proof. For fixed n ∈ N , setting √ 12 1 ln en √ fn (x) = , x ∈ [1, ∞), x ln enx ln ex √ 1 we find fn (1) = ln1en (2 ln en) 2 , and √ 12 √ 12 √ 1 1 ln en 1 ln en 1 (ln en) 2 0 √ √ , fn (x) = − 2 − 2 2 − 2 · √ x ln enx ln ex 2x ln enx (ln ex) 32 x ln enx ln ex √ 1 2 1 0 fn (1) = − + 2 (2 ln en) 2 . ln en ln en J. Inequal. Pure and Appl. Math., 3(5) Art. 75, 2002 http://jipam.vu.edu.au/ A N EW I NEQUALITY S IMILAR TO H ILBERT ’ S I NEQUALITY √ ln ex √ ln en Setting u = in the following integral, we obtain ∞ Z 3 ∞ Z fn (x)dx = 1 2 ln 1√ en 1 1+u 12 12 Z 1√ 2 ln en 1 1 1 du. du = π − 1+u u u 0 Hence by (2.1), (2.2) and Lemma 2.1, we have ω(n) = ∞ X ∞ Z 1 1 fn (x)dx + fn (1) − fn0 (1) 2 12 12 √ 1 2 1 1 1 du + + (2 ln en) 2 2 3 ln en 12 ln en 1+u u fn (m) < 1 m=1 Z 1/(2 ln √ en) =π− 0 = π − (2 ln √ 1 en) 2 R(n) < π. The lemma is proved. Lemma 2.3. For 0 < < 1, we have (2.3) ∞ X ∞ X 1 mn ln emn n=1 m=1 Proof. Setting u = Z ∞ √ e 1 x ln exy √ ln ex √ ln ey 1 √ 1 √ √ ln em ln en 1+ 2 1 > (π + o(1)) ( → 0+ ). in the following integral, we find 1+ 2 ln ex 1+ 2 1 √ = ln ey 1+ 2 1 √ = ln ey 1+ 2 1 √ > ln ey 1+ 2 1 √ = ln ey dx 1+ 1 1 2 du 1 1+u u √ ln ey 1+ 1+ 1+ Z √1 Z ∞ 2 ln ey 1 1 1 2 1 1 2 √ du − du 1+u u 1+u u ln ey 0 0 1+ 1+ Z √1 1+ Z ∞ 2 ln ey 1 2 1 1 2 1 √ du − du 1+u u u ln ey 0 0 2 1 √ (π + o(1)) − ( −→ 0+ ). 1 − ln ey Z ∞ Hence we have 1+ 2 1 1 √ √ mn ln emn ln em ln en n=1 m=1 1+ Z ∞Z ∞ 2 1 1 √ √ dxdy > √ √ ln ex ln ey e e xy ln exy # "Z 1+ 1+ Z ∞ ∞ 2 2 1 1 1 1 √ √ dx dy = √ √ x ln exy ln ex ln ey e e y ∞ X ∞ X J. Inequal. Pure and Appl. Math., 3(5) Art. 75, 2002 http://jipam.vu.edu.au/ 4 B ICHENG YANG Z > (π + o(1)) ∞ √ 1 y 1 √ 1+ 2 dy − 1− Z ∞ √ ln ey e 4 1 1 = (π + o(1)) − = (π + o(1)) ( → 0+ ). 1 − 2 e 1 y 1 √ 1+ +1 2 ln ey dy The lemma is proved. 3. M AIN R ESULT AND AN A PPLICATION P∞ P 2 Theorem 3.1. If 0 < n=1 na2n < ∞ and 0 < ∞ n=1 nbn < ∞, then ! 12 ∞ X ∞ ∞ ∞ X X X am bn (3.1) , <π na2n nb2n ln emn n=1 m=1 n=1 n=1 where the constant factor π is the best possible. Proof. By Cauchy’s inequality and (1.3), we have ∞ X ∞ X am b n ln emn n=1 m=1 #" √ 14 1 # √ 1 ln em 4 m 12 ln en n 2 bn √ √ = 1 1 n m ln en 2 (ln emn) 2 ln em n=1 m=1 (ln emn) " ∞ ∞ 1 √ 12 # 12 ∞ X ∞ 2 X X a2 ln √em 2 m X bn ln en n m √ √ ≤ ln emn ln en n n=1 m=1 ln emn ln em m m=1 n=1 1 ! ∞ ∞ 2 X X 2 2 = ω(m)mam ω(n)nbn . ∞ X ∞ X m=1 " am n=1 By Lemma 2.2, we have (3.1). For 0 < < 1, setting a0n as: a0n = then we have ∞ X n=1 (3.2) 1 √ 1+ , n ∈ N, n(ln en) 2 ∞ na0n 2 X 1 1 1 √ 1+ + √ 1+ + √ 1+ = (ln e) 2(ln 2 e) n(ln en) n=3 Z ∞ 1 1 1 √ 1+ dx √ 1+ + √ 1+ + √ < (ln e) 2(ln 2 e) e x(ln ex) 1 1 1 1 √ 1+ + √ 1+ + = (1 + o(1)) ( → 0+ ). = (ln e) 2(ln 2 e) If the constant factor π in (3.1) is not the best possible, then there exists a positive number K < π, such that (3.1) is valid if we change π to K. In particular, we have ∞ X ∞ ∞ X X a0m a0n <K na0n 2 . ln emn n=1 m=1 n=1 J. Inequal. Pure and Appl. Math., 3(5) Art. 75, 2002 http://jipam.vu.edu.au/ A N EW I NEQUALITY S IMILAR TO H ILBERT ’ S I NEQUALITY 5 By (2.3) and (3.2), we have ∞ X ∞ X a0m a0n (π + o(1)) < < K(1 + o(1)) ( → 0+ ), ln emn n=1 m=1 and π ≤ K. This contradicts that K < π. Hence the constant factor π in (3.1) is the best possible. The theorem is proved. Remark 3.2. Inequality (3.1) is more similar to the following Mulholland’s inequality for p = q = 2 (see [7]): ! p1 ∞ ! 1q ∞ ∞ ∞ X X X X π am b n . < n−1 apn n−1 bqn (3.3) π mn ln emn sin( ) p n=2 n=2 n=2 m=2 P 2 Theorem 3.3. If 0 < ∞ n=1 nan < ∞, then we have !2 ∞ ∞ ∞ X X X 1 am 2 (3.4) <π na2n , n m=1 ln emn n=1 n=1 where the constant factor π 2 is the best possible. Inequalities (3.1) and (3.4) are equivalent. P P Proof. Since ∞ na2n > 0, there exists k0 ≥ 1, such that for any k > k0 , we have kn=1 na2n > n=1 P |am | 0, and bn (k) = n1 km=1 lnemn > 0 (n ∈ N ). By (3.1), we have " k #2 X 0< nb2n (k) n=1 k X 1 = n n=1 " (3.5) k X |am | ln emn m=1 k X k X |am |bn (k) = ln emn n=1 m=1 !2 2 #2 <π 2 k X n=1 na2n k X nb2n (k). n=1 Thus we find !2 k k k k X X X 1 X |am | 2 2 (3.6) 0< = nbn (k) < π na2n . n ln emn n=1 m=1 n=1 n=1 P∞ P∞ 2 2 2 It follows that 0 < n=1 nbn (∞) ≤ π n=1 nan < ∞. Hence by (3.1), for k → ∞, neither (3.5) nor (3.6) takes equality, and we have !2 !2 ∞ ∞ ∞ ∞ ∞ X X X 1 X am 1 X |am | 2 ≤ <π na2n . n ln emn n ln emn n=1 m=1 n=1 m=1 n=1 Inequality (3.4) is valid. On the other hand, if (3.4) holds, by Cauchy’s inequality, we have ! ∞ X ∞ ∞ ∞ 1 X X am b n 1 X am n 2 bn = 1 ln emn ln emn 2 n n=1 m=1 m=1 n=1 21 !2 ∞ ∞ ∞ X X X 1 am (3.7) ≤ nb2n . n ln emn n=1 m=1 n=1 J. Inequal. Pure and Appl. Math., 3(5) Art. 75, 2002 http://jipam.vu.edu.au/ 6 B ICHENG YANG By (3.4), we have (3.1). Hence inequalities (3.1) and (3.4) are equivalent. If the constant factor π 2 in (3.4) is not the best possible, we may show that the constant factor π in (3.1) is not the best possible, by using (3.7). This is a contradiction. The theorem is proved. Remark 3.4. Inequality (3.4) is similar to the following equivalent form of (1.1) (see [2]): !2 ∞ ∞ ∞ X X X am (3.8) < π2 a2n . m + n + 1 m=0 n=0 n=0 Since inequalities (3.1) and (3.4) are similar to (1.1) and its equivalent form with the best constant factors, we have provided some new results. R EFERENCES [1] G.H. HARDY, J.E. LITTLEWOOD AND G. POLYA, Inequalities, Cambridge Univ. Press, London, 1952. [2] B. YANG AND L. DEBNATH, on a new generalization of Hardy-Hilbert’s inequality and its application, J. Math. Anal. Appl., 233 (1999), 484–497. [3] B. YANG AND L. DEBNATH, Some inequalities involving π and an application to Hilbert’s inequality, Applied Math. Letters, 129 (1999), 101–105. [4] B. YANG, On a strengthened version of the more accurate Hardy-Hilbert’s inequality, Acta Math. Sinica, 42(6) (1999), 1103–1110. [5] B. YANG, On a strengthened Hardy-Hilbert’s inequality, J. Ineq. Pure. and Appl. Math., 1(2), Art. 22 (2000). [ONLINE: http://jipam.vu.edu.au/v1n2/012_00.html] [6] J. KUANG AND L. DEBNATH, On new generalizations of Hilbert’s inequality and their applications, J. Math. Anal. Appl., 245 (2000), 248–265. [7] H.P. MULHOLLAND, Some theorem on Dirichlet series with coefficients and related integrals, Proc. London Math. Soc., 29(2) (1999), 281–292. J. Inequal. Pure and Appl. Math., 3(5) Art. 75, 2002 http://jipam.vu.edu.au/