Text-overlap comparison
8 candidate passages (embedding cosine similarity 0.91–1.00) identified this pair for review. The exhibits below are exact word-shingle matches instead — a multi-word verbatim run is the more convincing evidence of shared text than embedding similarity, with the matched span highlighted.
EARLIER¶0-4
Lai Journal of Inequalities and Applications ( 2015) 2015:357 DOI 10.1186/s13660-015-0874-2 R ES EA RCH Open Access The integer part of a nonlinear form with integer variables Kai Lai* *Correspondence: k.ching.lai@foxmail.com College of Computer and Information Engineering, Henan University of Economics and Law, Zhengzhou, 450011, P.R. China Abstract Using the Davenport-Heilbronn method, we show that if λ1,λ2,...,λ9 are positive real numbers, at least one of the ratios λi/λj (1 ≤ i < j ≤ 9) is irrational, then the integer parts of λ1x3 1 + λ2x3 2 + λ3x4 3 + λ4x4 4 + λ5x5 5 + ··· + λ9x5 9 are prime infinitely often for natural numbers x1,x2,...,x9. Keywords: Davenport-Heilbronn method; integer variables; diophantine approximation 1 Introduction In , Brüdern et al. [] proved that if λ,...,λs are positive real numbers, λ/λ is irrational, all Dirichlet L-functions satisfy the Riemann hypothesis s ≥ k +, then the integer parts of λxk + λxk + ··· + λsxk s are prime infinitely often for natural numbers xj. Motivated by [], using the Davenport-Heilbronn method, we consider the integer part of a nonlinear form with integer variables and mixed powers , and , and establish one result as follows.
LATER¶0-4
Wang and Mu Journal of Inequalities and Applications ( 2017) 2017:170 DOI 10.1186/s13660-017-1440-x R ES EA RCH Open Access An extension of the mixed integer part of a nonlinear form Yunhan Wang1 and Jiani Mu2* *Correspondence: 3289807821@qq.com 2Faculty of Science and Technology, Tapee College, Surathani, 84000, Thailand Full list of author information is available at the end of the article Abstract Our aim in this paper is to consider the integer part of a nonlinear form representing primes. We establish that if λ1,λ2,...,λ8 are positive real numbers, at least one of the ratios λi/λj (1 ≤ i < j ≤ 8) is irrational, then the integer parts of λ1x3 1 + λ2x3 2 + λ3x4 3 + λ4x4 4 + λ5x4 5 + λ6x5 6 + λ7x5 7 + λ8x5 8 are prime infinitely often for x1,x2,...,x8, where x1,x2,...,x8 are natural numbers. Keywords: diophantine approximation; mixed power; integer variables 1 Introduction The integer part of linear and nonlinear forms representing primes has been considered by many scholars. Let [x] be the greatest integer not exceeding x. In , Danicic [] proved that if the diophantine inequality |λp + λp + λp + η| < ε () satisfies certain conditions, and primes pi ≤ N (i = ,,), then the number of prime solutions (p,p,p,p) of () is greater than CN(logN)–, where C is a positive number independent of N. Based on the above result, Danicic [] proved that if λ, μ are non-zero real numbers, not both negative, λ is irrational, and m is a positive integer, then there exist infinitely many primes p and pairs of primes p, p and p such that
EARLIER¶0-4
Lai Journal of Inequalities and Applications ( 2015) 2015:357 DOI 10.1186/s13660-015-0874-2 R ES EA RCH Open Access The integer part of a nonlinear form with integer variables Kai Lai* *Correspondence: k.ching.lai@foxmail.com College of Computer and Information Engineering, Henan University of Economics and Law, Zhengzhou, 450011, P.R. China Abstract Using the Davenport-Heilbronn method, we show that if λ1,λ2,...,λ9 are positive real numbers, at least one of the ratios λi/λj (1 ≤ i < j ≤ 9) is irrational, then the integer parts of λ1x3 1 + λ2x3 2 + λ3x4 3 + λ4x4 4 + λ5x5 5 + ··· + λ9x5 9 are prime infinitely often for natural numbers x1,x2,...,x9. Keywords: Davenport-Heilbronn method; integer variables; diophantine approximation 1 Introduction In , Brüdern et al. [] proved that if λ,...,λs are positive real numbers, λ/λ is irrational, all Dirichlet L-functions satisfy the Riemann hypothesis s ≥ k +, then the integer parts of λxk + λxk + ··· + λsxk s are prime infinitely often for natural numbers xj. Motivated by [], using the Davenport-Heilbronn method, we consider the integer part of a nonlinear form with integer variables and mixed powers , and , and establish one result as follows.
LATER¶0-4
Wang and Mu Journal of Inequalities and Applications ( 2017) 2017:170 DOI 10.1186/s13660-017-1440-x R ES EA RCH Open Access An extension of the mixed integer part of a nonlinear form Yunhan Wang1 and Jiani Mu2* *Correspondence: 3289807821@qq.com 2Faculty of Science and Technology, Tapee College, Surathani, 84000, Thailand Full list of author information is available at the end of the article Abstract Our aim in this paper is to consider the integer part of a nonlinear form representing primes. We establish that if λ1,λ2,...,λ8 are positive real numbers, at least one of the ratios λi/λj (1 ≤ i < j ≤ 8) is irrational, then the integer parts of λ1x3 1 + λ2x3 2 + λ3x4 3 + λ4x4 4 + λ5x4 5 + λ6x5 6 + λ7x5 7 + λ8x5 8 are prime infinitely often for x1,x2,...,x8, where x1,x2,...,x8 are natural numbers. Keywords: diophantine approximation; mixed power; integer variables 1 Introduction The integer part of linear and nonlinear forms representing primes has been considered by many scholars. Let [x] be the greatest integer not exceeding x. In , Danicic [] proved that if the diophantine inequality |λp + λp + λp + η| < ε () satisfies certain conditions, and primes pi ≤ N (i = ,,), then the number of prime solutions (p,p,p,p) of () is greater than CN(logN)–, where C is a positive number independent of N. Based on the above result, Danicic [] proved that if λ, μ are non-zero real numbers, not both negative, λ is irrational, and m is a positive integer, then there exist infinitely many primes p and pairs of primes p, p and p such that
EARLIER¶1-5
Kai Lai* *Correspondence: k.ching.lai@foxmail.com College of Computer and Information Engineering, Henan University of Economics and Law, Zhengzhou, 450011, P.R. China Abstract Using the Davenport-Heilbronn method, we show that if λ1,λ2,...,λ9 are positive real numbers, at least one of the ratios λi/λj (1 ≤ i < j ≤ 9) is irrational, then the integer parts of λ1x3 1 + λ2x3 2 + λ3x4 3 + λ4x4 4 + λ5x5 5 + ··· + λ9x5 9 are prime infinitely often for natural numbers x1,x2,...,x9. Keywords: Davenport-Heilbronn method; integer variables; diophantine approximation 1 Introduction In , Brüdern et al. [] proved that if λ,...,λs are positive real numbers, λ/λ is irrational, all Dirichlet L-functions satisfy the Riemann hypothesis s ≥ k +, then the integer parts of λxk + λxk + ··· + λsxk s are prime infinitely often for natural numbers xj. Motivated by [], using the Davenport-Heilbronn method, we consider the integer part of a nonlinear form with integer variables and mixed powers , and , and establish one result as follows. Theorem . Let λ,λ,...,λ be positive real numbers, at least one of the ratios λi/λj ( ≤ i < j ≤ ) is irrational. Then the integer parts of λx + λx + λx + λx + λx + ··· + λx
LATER¶3-7
Keywords: diophantine approximation; mixed power; integer variables 1 Introduction The integer part of linear and nonlinear forms representing primes has been considered by many scholars. Let [x] be the greatest integer not exceeding x. In , Danicic [] proved that if the diophantine inequality |λp + λp + λp + η| < ε () satisfies certain conditions, and primes pi ≤ N (i = ,,), then the number of prime solutions (p,p,p,p) of () is greater than CN(logN)–, where C is a positive number independent of N. Based on the above result, Danicic [] proved that if λ, μ are non-zero real numbers, not both negative, λ is irrational, and m is a positive integer, then there exist infinitely many primes p and pairs of primes p, p and p such that [λp + μp + μp] = mp. In particular, [λp + μp + μp] represents infinitely many primes. Brüdern et al. [] proved that if λ,...,λs are positive real numbers, λ/λ is irrational, all Dirichlet L-functions satisfy the Riemann hypothesis, s ≥ k + , then the integer parts of λxk + λxk + ··· + λsxk s are prime infinitely often for natural numbers xj, where xj is a natural number. Recently, Lai [] proved that for integer k ≥ , r ≥ k– + , under certain conditions, there exist infinitely many primes p,...,pr,p such that � μpk + ··· + μrpk r � = mp. (.) © The Author(s) 2017. This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.
EARLIER¶2-6
Abstract Using the Davenport-Heilbronn method, we show that if λ1,λ2,...,λ9 are positive real numbers, at least one of the ratios λi/λj (1 ≤ i < j ≤ 9) is irrational, then the integer parts of λ1x3 1 + λ2x3 2 + λ3x4 3 + λ4x4 4 + λ5x5 5 + ··· + λ9x5 9 are prime infinitely often for natural numbers x1,x2,...,x9. Keywords: Davenport-Heilbronn method; integer variables; diophantine approximation 1 Introduction In , Brüdern et al. [] proved that if λ,...,λs are positive real numbers, λ/λ is irrational, all Dirichlet L-functions satisfy the Riemann hypothesis s ≥ k +, then the integer parts of λxk + λxk + ··· + λsxk s are prime infinitely often for natural numbers xj. Motivated by [], using the Davenport-Heilbronn method, we consider the integer part of a nonlinear form with integer variables and mixed powers , and , and establish one result as follows. Theorem . Let λ,λ,...,λ be positive real numbers, at least one of the ratios λi/λj ( ≤ i < j ≤ ) is irrational. Then the integer parts of λx + λx + λx + λx + λx + ··· + λx are prime infinitely often for natural numbers x,x,...,x. It is noted that Theorem . holds without the Riemann hypothesis. 2 Notation Throughout, we use p to denote a prime number and xj to denote a natural number. We denote by δ a sufficiently small positive number and by ε an arbitrarily small positive number. Constants, both explicit and implicit, in Landau or Vinogradov symbols may depend
LATER¶4-8
|λp + λp + λp + η| < ε () satisfies certain conditions, and primes pi ≤ N (i = ,,), then the number of prime solutions (p,p,p,p) of () is greater than CN(logN)–, where C is a positive number independent of N. Based on the above result, Danicic [] proved that if λ, μ are non-zero real numbers, not both negative, λ is irrational, and m is a positive integer, then there exist infinitely many primes p and pairs of primes p, p and p such that [λp + μp + μp] = mp. In particular, [λp + μp + μp] represents infinitely many primes. Brüdern et al. [] proved that if λ,...,λs are positive real numbers, λ/λ is irrational, all Dirichlet L-functions satisfy the Riemann hypothesis, s ≥ k + , then the integer parts of λxk + λxk + ··· + λsxk s are prime infinitely often for natural numbers xj, where xj is a natural number. Recently, Lai [] proved that for integer k ≥ , r ≥ k– + , under certain conditions, there exist infinitely many primes p,...,pr,p such that � μpk + ··· + μrpk r � = mp. (.) © The Author(s) 2017. This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. Wang and Mu Journal of Inequalities and Applications ( 2017) 2017:170 Page 2 of 9 It is natural to ask if the above results are true when primes pj in (.) are replaced by natural numbers xj. In this paper we shall give an affirmative answer to this question.
EARLIER¶5-9
Theorem . Let λ,λ,...,λ be positive real numbers, at least one of the ratios λi/λj ( ≤ i < j ≤ ) is irrational. Then the integer parts of λx + λx + λx + λx + λx + ··· + λx are prime infinitely often for natural numbers x,x,...,x. It is noted that Theorem . holds without the Riemann hypothesis. 2 Notation Throughout, we use p to denote a prime number and xj to denote a natural number. We denote by δ a sufficiently small positive number and by ε an arbitrarily small positive number. Constants, both explicit and implicit, in Landau or Vinogradov symbols may depend © 2015 Lai. This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. Lai Journal of Inequalities and Applications ( 2015) 2015:357 Page 2 of 9 on λ,λ,...,λ. We write e(x) = exp(πix). We use [x] to denote the integer part of real variable x. We take X to be the basic parameter, a large real integer. Since at least one of the ratios λi/λj ( ≤ i < j ≤ ) is irrational, without loss of generality we may assume that
LATER¶5-9
[λp + μp + μp] = mp. In particular, [λp + μp + μp] represents infinitely many primes. Brüdern et al. [] proved that if λ,...,λs are positive real numbers, λ/λ is irrational, all Dirichlet L-functions satisfy the Riemann hypothesis, s ≥ k + , then the integer parts of λxk + λxk + ··· + λsxk s are prime infinitely often for natural numbers xj, where xj is a natural number. Recently, Lai [] proved that for integer k ≥ , r ≥ k– + , under certain conditions, there exist infinitely many primes p,...,pr,p such that � μpk + ··· + μrpk r � = mp. (.) © The Author(s) 2017. This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. Wang and Mu Journal of Inequalities and Applications ( 2017) 2017:170 Page 2 of 9 It is natural to ask if the above results are true when primes pj in (.) are replaced by natural numbers xj. In this paper we shall give an affirmative answer to this question. 2 Main result Our main aim is to investigate the integer part of a nonlinear form with integer variables and mixed powers , and . Using Tumura-Clunie type inequalities (see [, ]), we establish one result as follows.
EARLIER¶2-6
Abstract Using the Davenport-Heilbronn method, we show that if λ1,λ2,...,λ9 are positive real numbers, at least one of the ratios λi/λj (1 ≤ i < j ≤ 9) is irrational, then the integer parts of λ1x3 1 + λ2x3 2 + λ3x4 3 + λ4x4 4 + λ5x5 5 + ··· + λ9x5 9 are prime infinitely often for natural numbers x1,x2,...,x9. Keywords: Davenport-Heilbronn method; integer variables; diophantine approximation 1 Introduction In , Brüdern et al. [] proved that if λ,...,λs are positive real numbers, λ/λ is irrational, all Dirichlet L-functions satisfy the Riemann hypothesis s ≥ k +, then the integer parts of λxk + λxk + ··· + λsxk s are prime infinitely often for natural numbers xj. Motivated by [], using the Davenport-Heilbronn method, we consider the integer part of a nonlinear form with integer variables and mixed powers , and , and establish one result as follows. Theorem . Let λ,λ,...,λ be positive real numbers, at least one of the ratios λi/λj ( ≤ i < j ≤ ) is irrational. Then the integer parts of λx + λx + λx + λx + λx + ··· + λx are prime infinitely often for natural numbers x,x,...,x. It is noted that Theorem . holds without the Riemann hypothesis. 2 Notation Throughout, we use p to denote a prime number and xj to denote a natural number. We denote by δ a sufficiently small positive number and by ε an arbitrarily small positive number. Constants, both explicit and implicit, in Landau or Vinogradov symbols may depend
LATER¶7-11
� μpk + ··· + μrpk r � = mp. (.) © The Author(s) 2017. This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. Wang and Mu Journal of Inequalities and Applications ( 2017) 2017:170 Page 2 of 9 It is natural to ask if the above results are true when primes pj in (.) are replaced by natural numbers xj. In this paper we shall give an affirmative answer to this question. 2 Main result Our main aim is to investigate the integer part of a nonlinear form with integer variables and mixed powers , and . Using Tumura-Clunie type inequalities (see [, ]), we establish one result as follows. Theorem . Let λ,λ,...,λ be nonnegative real numbers, at least one of the ratios λi/λj ( ≤ i < j ≤ ) is rational. Then the integer parts of λx + λx + λx + λx + λx + λx + λx + λx are prime infinitely often for x,x,...,x, where x,x,...,x are natural numbers. Remark It is easy to see from Theorem . that primes pj in (.) are replaced by natural numbers xj and there exist infinitely many primes p,...,pr and p such that [μpk + ··· + μr+pk r+] = mpr, where m is a nonnegative integer (see []).
EARLIER¶3-7
Keywords: Davenport-Heilbronn method; integer variables; diophantine approximation 1 Introduction In , Brüdern et al. [] proved that if λ,...,λs are positive real numbers, λ/λ is irrational, all Dirichlet L-functions satisfy the Riemann hypothesis s ≥ k +, then the integer parts of λxk + λxk + ··· + λsxk s are prime infinitely often for natural numbers xj. Motivated by [], using the Davenport-Heilbronn method, we consider the integer part of a nonlinear form with integer variables and mixed powers , and , and establish one result as follows. Theorem . Let λ,λ,...,λ be positive real numbers, at least one of the ratios λi/λj ( ≤ i < j ≤ ) is irrational. Then the integer parts of λx + λx + λx + λx + λx + ··· + λx are prime infinitely often for natural numbers x,x,...,x. It is noted that Theorem . holds without the Riemann hypothesis. 2 Notation Throughout, we use p to denote a prime number and xj to denote a natural number. We denote by δ a sufficiently small positive number and by ε an arbitrarily small positive number. Constants, both explicit and implicit, in Landau or Vinogradov symbols may depend © 2015 Lai. This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.
LATER¶8-12
Wang and Mu Journal of Inequalities and Applications ( 2017) 2017:170 Page 2 of 9 It is natural to ask if the above results are true when primes pj in (.) are replaced by natural numbers xj. In this paper we shall give an affirmative answer to this question. 2 Main result Our main aim is to investigate the integer part of a nonlinear form with integer variables and mixed powers , and . Using Tumura-Clunie type inequalities (see [, ]), we establish one result as follows. Theorem . Let λ,λ,...,λ be nonnegative real numbers, at least one of the ratios λi/λj ( ≤ i < j ≤ ) is rational. Then the integer parts of λx + λx + λx + λx + λx + λx + λx + λx are prime infinitely often for x,x,...,x, where x,x,...,x are natural numbers. Remark It is easy to see from Theorem . that primes pj in (.) are replaced by natural numbers xj and there exist infinitely many primes p,...,pr and p such that [μpk + ··· + μr+pk r+] = mpr, where m is a nonnegative integer (see []). 3 Outline of the proof Throughout this paper, p denotes a prime number, and xj denotes a natural number. δ is a sufficiently small positive number, ε is an arbitrarily small positive number. Constants, both explicit and implicit, in Landau or Vinogradov symbols may depend on λ,λ,...,λ. We write e(x) = exp(πix). We take X to be the basic parameter, a large real integer. Since at least one of the ratios λi/λj ( ≤ i < j ≤ ) is irrational, without loss of generality, we may assume that λ/λ is irrational. For the other cases, the only difference is in the following intermediate region, and we may deal with the same method in Section . Since λ/λ is irrational, there are infinitely many pairs of integers q, a with |λ/λ – a/q| ≥ q–, (p,q) = , q > and a ̸= . We choose p to be large in terms of λ,λ,...,λ, and make the following definitions. Put τ = N–+δ,T = N , L = logN, Q = (|λ|– + |λ|–)N–δ, [N–δ] = p and P = Nδ, where N ≍ X. Let ν be a positive real number, we define
EARLIER¶7-11
© 2015 Lai. This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. Lai Journal of Inequalities and Applications ( 2015) 2015:357 Page 2 of 9 on λ,λ,...,λ. We write e(x) = exp(πix). We use [x] to denote the integer part of real variable x. We take X to be the basic parameter, a large real integer. Since at least one of the ratios λi/λj ( ≤ i < j ≤ ) is irrational, without loss of generality we may assume that λ/λ is irrational. For the other cases, the only difference is in the following intermediate region, and we may deal with the same method in Section . Since λ/λ is irrational, then there are infinitely many pairs of integers q, a with |λ/λ – a/q| ≤ q–, (a,q) = , q > and a ̸= . We choose q to be large in terms of λ,λ,...,λ and
LATER¶8-12
Wang and Mu Journal of Inequalities and Applications ( 2017) 2017:170 Page 2 of 9 It is natural to ask if the above results are true when primes pj in (.) are replaced by natural numbers xj. In this paper we shall give an affirmative answer to this question. 2 Main result Our main aim is to investigate the integer part of a nonlinear form with integer variables and mixed powers , and . Using Tumura-Clunie type inequalities (see [, ]), we establish one result as follows. Theorem . Let λ,λ,...,λ be nonnegative real numbers, at least one of the ratios λi/λj ( ≤ i < j ≤ ) is rational. Then the integer parts of λx + λx + λx + λx + λx + λx + λx + λx are prime infinitely often for x,x,...,x, where x,x,...,x are natural numbers. Remark It is easy to see from Theorem . that primes pj in (.) are replaced by natural numbers xj and there exist infinitely many primes p,...,pr and p such that [μpk + ··· + μr+pk r+] = mpr, where m is a nonnegative integer (see []). 3 Outline of the proof Throughout this paper, p denotes a prime number, and xj denotes a natural number. δ is a sufficiently small positive number, ε is an arbitrarily small positive number. Constants, both explicit and implicit, in Landau or Vinogradov symbols may depend on λ,λ,...,λ. We write e(x) = exp(πix). We take X to be the basic parameter, a large real integer. Since at least one of the ratios λi/λj ( ≤ i < j ≤ ) is irrational, without loss of generality, we may assume that λ/λ is irrational. For the other cases, the only difference is in the following intermediate region, and we may deal with the same method in Section . Since λ/λ is irrational, there are infinitely many pairs of integers q, a with |λ/λ – a/q| ≥ q–, (p,q) = , q > and a ̸= . We choose p to be large in terms of λ,λ,...,λ, and make the following definitions. Put τ = N–+δ,T = N , L = logN, Q = (|λ|– + |λ|–)N–δ, [N–δ] = p and P = Nδ, where N ≍ X. Let ν be a positive real number, we define
EARLIER¶3-7
Keywords: Davenport-Heilbronn method; integer variables; diophantine approximation 1 Introduction In , Brüdern et al. [] proved that if λ,...,λs are positive real numbers, λ/λ is irrational, all Dirichlet L-functions satisfy the Riemann hypothesis s ≥ k +, then the integer parts of λxk + λxk + ··· + λsxk s are prime infinitely often for natural numbers xj. Motivated by [], using the Davenport-Heilbronn method, we consider the integer part of a nonlinear form with integer variables and mixed powers , and , and establish one result as follows. Theorem . Let λ,λ,...,λ be positive real numbers, at least one of the ratios λi/λj ( ≤ i < j ≤ ) is irrational. Then the integer parts of λx + λx + λx + λx + λx + ··· + λx are prime infinitely often for natural numbers x,x,...,x. It is noted that Theorem . holds without the Riemann hypothesis. 2 Notation Throughout, we use p to denote a prime number and xj to denote a natural number. We denote by δ a sufficiently small positive number and by ε an arbitrarily small positive number. Constants, both explicit and implicit, in Landau or Vinogradov symbols may depend © 2015 Lai. This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.
LATER¶8-12
Wang and Mu Journal of Inequalities and Applications ( 2017) 2017:170 Page 2 of 9 It is natural to ask if the above results are true when primes pj in (.) are replaced by natural numbers xj. In this paper we shall give an affirmative answer to this question. 2 Main result Our main aim is to investigate the integer part of a nonlinear form with integer variables and mixed powers , and . Using Tumura-Clunie type inequalities (see [, ]), we establish one result as follows. Theorem . Let λ,λ,...,λ be nonnegative real numbers, at least one of the ratios λi/λj ( ≤ i < j ≤ ) is rational. Then the integer parts of λx + λx + λx + λx + λx + λx + λx + λx are prime infinitely often for x,x,...,x, where x,x,...,x are natural numbers. Remark It is easy to see from Theorem . that primes pj in (.) are replaced by natural numbers xj and there exist infinitely many primes p,...,pr and p such that [μpk + ··· + μr+pk r+] = mpr, where m is a nonnegative integer (see []). 3 Outline of the proof Throughout this paper, p denotes a prime number, and xj denotes a natural number. δ is a sufficiently small positive number, ε is an arbitrarily small positive number. Constants, both explicit and implicit, in Landau or Vinogradov symbols may depend on λ,λ,...,λ. We write e(x) = exp(πix). We take X to be the basic parameter, a large real integer. Since at least one of the ratios λi/λj ( ≤ i < j ≤ ) is irrational, without loss of generality, we may assume that λ/λ is irrational. For the other cases, the only difference is in the following intermediate region, and we may deal with the same method in Section . Since λ/λ is irrational, there are infinitely many pairs of integers q, a with |λ/λ – a/q| ≥ q–, (p,q) = , q > and a ̸= . We choose p to be large in terms of λ,λ,...,λ, and make the following definitions. Put τ = N–+δ,T = N , L = logN, Q = (|λ|– + |λ|–)N–δ, [N–δ] = p and P = Nδ, where N ≍ X. Let ν be a positive real number, we define
EARLIER¶4-8
k +, then the integer parts of λxk + λxk + ··· + λsxk s are prime infinitely often for natural numbers xj. Motivated by [], using the Davenport-Heilbronn method, we consider the integer part of a nonlinear form with integer variables and mixed powers , and , and establish one result as follows. Theorem . Let λ,λ,...,λ be positive real numbers, at least one of the ratios λi/λj ( ≤ i < j ≤ ) is irrational. Then the integer parts of λx + λx + λx + λx + λx + ··· + λx are prime infinitely often for natural numbers x,x,...,x. It is noted that Theorem . holds without the Riemann hypothesis. 2 Notation Throughout, we use p to denote a prime number and xj to denote a natural number. We denote by δ a sufficiently small positive number and by ε an arbitrarily small positive number. Constants, both explicit and implicit, in Landau or Vinogradov symbols may depend © 2015 Lai. This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. Lai Journal of Inequalities and Applications ( 2015) 2015:357 Page 2 of 9 on λ,λ,...,λ. We write e(x) = exp(πix). We use [x] to denote the integer part of real
LATER¶10-14
Theorem . Let λ,λ,...,λ be nonnegative real numbers, at least one of the ratios λi/λj ( ≤ i < j ≤ ) is rational. Then the integer parts of λx + λx + λx + λx + λx + λx + λx + λx are prime infinitely often for x,x,...,x, where x,x,...,x are natural numbers. Remark It is easy to see from Theorem . that primes pj in (.) are replaced by natural numbers xj and there exist infinitely many primes p,...,pr and p such that [μpk + ··· + μr+pk r+] = mpr, where m is a nonnegative integer (see []). 3 Outline of the proof Throughout this paper, p denotes a prime number, and xj denotes a natural number. δ is a sufficiently small positive number, ε is an arbitrarily small positive number. Constants, both explicit and implicit, in Landau or Vinogradov symbols may depend on λ,λ,...,λ. We write e(x) = exp(πix). We take X to be the basic parameter, a large real integer. Since at least one of the ratios λi/λj ( ≤ i < j ≤ ) is irrational, without loss of generality, we may assume that λ/λ is irrational. For the other cases, the only difference is in the following intermediate region, and we may deal with the same method in Section . Since λ/λ is irrational, there are infinitely many pairs of integers q, a with |λ/λ – a/q| ≥ q–, (p,q) = , q > and a ̸= . We choose p to be large in terms of λ,λ,...,λ, and make the following definitions. Put τ = N–+δ,T = N , L = logN, Q = (|λ|– + |λ|–)N–δ, [N–δ] = p and P = Nδ, where N ≍ X. Let ν be a positive real number, we define e(αx)dx. (.) Wang and Mu Journal of Inequalities and Applications ( 2017) 2017:170 Page 3 of 9 From (.) we have J =: � +∞ –∞ � i= Fi(λiα)G(–α)e � – α � K (α)dα ≤ logN � |λx +λx +λx +λx +λx +···+λx –p– |< ≤x,x≤X/,≤x,x≤X/,≤x,...,x≤X/,p≤N , which gives that
EARLIER¶7-11
© 2015 Lai. This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. Lai Journal of Inequalities and Applications ( 2015) 2015:357 Page 2 of 9 on λ,λ,...,λ. We write e(x) = exp(πix). We use [x] to denote the integer part of real variable x. We take X to be the basic parameter, a large real integer. Since at least one of the ratios λi/λj ( ≤ i < j ≤ ) is irrational, without loss of generality we may assume that λ/λ is irrational. For the other cases, the only difference is in the following intermediate region, and we may deal with the same method in Section . Since λ/λ is irrational, then there are infinitely many pairs of integers q, a with |λ/λ – a/q| ≤ q–, (a,q) = , q > and a ̸= . We choose q to be large in terms of λ,λ,...,λ and
LATER¶10-14
Theorem . Let λ,λ,...,λ be nonnegative real numbers, at least one of the ratios λi/λj ( ≤ i < j ≤ ) is rational. Then the integer parts of λx + λx + λx + λx + λx + λx + λx + λx are prime infinitely often for x,x,...,x, where x,x,...,x are natural numbers. Remark It is easy to see from Theorem . that primes pj in (.) are replaced by natural numbers xj and there exist infinitely many primes p,...,pr and p such that [μpk + ··· + μr+pk r+] = mpr, where m is a nonnegative integer (see []). 3 Outline of the proof Throughout this paper, p denotes a prime number, and xj denotes a natural number. δ is a sufficiently small positive number, ε is an arbitrarily small positive number. Constants, both explicit and implicit, in Landau or Vinogradov symbols may depend on λ,λ,...,λ. We write e(x) = exp(πix). We take X to be the basic parameter, a large real integer. Since at least one of the ratios λi/λj ( ≤ i < j ≤ ) is irrational, without loss of generality, we may assume that λ/λ is irrational. For the other cases, the only difference is in the following intermediate region, and we may deal with the same method in Section . Since λ/λ is irrational, there are infinitely many pairs of integers q, a with |λ/λ – a/q| ≥ q–, (p,q) = , q > and a ̸= . We choose p to be large in terms of λ,λ,...,λ, and make the following definitions. Put τ = N–+δ,T = N , L = logN, Q = (|λ|– + |λ|–)N–δ, [N–δ] = p and P = Nδ, where N ≍ X. Let ν be a positive real number, we define e(αx)dx. (.) Wang and Mu Journal of Inequalities and Applications ( 2017) 2017:170 Page 3 of 9 From (.) we have J =: � +∞ –∞ � i= Fi(λiα)G(–α)e � – α � K (α)dα ≤ logN � |λx +λx +λx +λx +λx +···+λx –p– |< ≤x,x≤X/,≤x,x≤X/,≤x,...,x≤X/,p≤N , which gives that
EARLIER¶3-7
Keywords: Davenport-Heilbronn method; integer variables; diophantine approximation 1 Introduction In , Brüdern et al. [] proved that if λ,...,λs are positive real numbers, λ/λ is irrational, all Dirichlet L-functions satisfy the Riemann hypothesis s ≥ k +, then the integer parts of λxk + λxk + ··· + λsxk s are prime infinitely often for natural numbers xj. Motivated by [], using the Davenport-Heilbronn method, we consider the integer part of a nonlinear form with integer variables and mixed powers , and , and establish one result as follows. Theorem . Let λ,λ,...,λ be positive real numbers, at least one of the ratios λi/λj ( ≤ i < j ≤ ) is irrational. Then the integer parts of λx + λx + λx + λx + λx + ··· + λx are prime infinitely often for natural numbers x,x,...,x. It is noted that Theorem . holds without the Riemann hypothesis. 2 Notation Throughout, we use p to denote a prime number and xj to denote a natural number. We denote by δ a sufficiently small positive number and by ε an arbitrarily small positive number. Constants, both explicit and implicit, in Landau or Vinogradov symbols may depend © 2015 Lai. This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.
LATER¶10-14
Theorem . Let λ,λ,...,λ be nonnegative real numbers, at least one of the ratios λi/λj ( ≤ i < j ≤ ) is rational. Then the integer parts of λx + λx + λx + λx + λx + λx + λx + λx are prime infinitely often for x,x,...,x, where x,x,...,x are natural numbers. Remark It is easy to see from Theorem . that primes pj in (.) are replaced by natural numbers xj and there exist infinitely many primes p,...,pr and p such that [μpk + ··· + μr+pk r+] = mpr, where m is a nonnegative integer (see []). 3 Outline of the proof Throughout this paper, p denotes a prime number, and xj denotes a natural number. δ is a sufficiently small positive number, ε is an arbitrarily small positive number. Constants, both explicit and implicit, in Landau or Vinogradov symbols may depend on λ,λ,...,λ. We write e(x) = exp(πix). We take X to be the basic parameter, a large real integer. Since at least one of the ratios λi/λj ( ≤ i < j ≤ ) is irrational, without loss of generality, we may assume that λ/λ is irrational. For the other cases, the only difference is in the following intermediate region, and we may deal with the same method in Section . Since λ/λ is irrational, there are infinitely many pairs of integers q, a with |λ/λ – a/q| ≥ q–, (p,q) = , q > and a ̸= . We choose p to be large in terms of λ,λ,...,λ, and make the following definitions. Put τ = N–+δ,T = N , L = logN, Q = (|λ|– + |λ|–)N–δ, [N–δ] = p and P = Nδ, where N ≍ X. Let ν be a positive real number, we define e(αx)dx. (.) Wang and Mu Journal of Inequalities and Applications ( 2017) 2017:170 Page 3 of 9 From (.) we have J =: � +∞ –∞ � i= Fi(λiα)G(–α)e � – α � K (α)dα ≤ logN � |λx +λx +λx +λx +λx +···+λx –p– |< ≤x,x≤X/,≤x,x≤X/,≤x,...,x≤X/,p≤N , which gives that
EARLIER¶7-13
© 2015 Lai. This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. Lai Journal of Inequalities and Applications ( 2015) 2015:357 Page 2 of 9 on λ,λ,...,λ. We write e(x) = exp(πix). We use [x] to denote the integer part of real variable x. We take X to be the basic parameter, a large real integer. Since at least one of the ratios λi/λj ( ≤ i < j ≤ ) is irrational, without loss of generality we may assume that λ/λ is irrational. For the other cases, the only difference is in the following intermediate region, and we may deal with the same method in Section . Since λ/λ is irrational, then there are infinitely many pairs of integers q, a with |λ/λ – a/q| ≤ q–, (a,q) = , q > and a ̸= . We choose q to be large in terms of λ,λ,...,λ and make the following definitions. N ≍ X, L = logN, � N–δ� = q, τ = N–+δ, Q = � |λ|– + |λ|–� N–δ, P = Nδ, T = N . Let ν be a positive real number, we define Kν(α) = ν �sinπνα πνα � , α ̸= , Kν() = ν, (.) Fi(α) = � ≤x≤X e � αx� , i = ,, Fj(α) = � ≤x≤X e � αx� , j = ,, Fk(α) = � ≤x≤X
LATER¶10-14
Theorem . Let λ,λ,...,λ be nonnegative real numbers, at least one of the ratios λi/λj ( ≤ i < j ≤ ) is rational. Then the integer parts of λx + λx + λx + λx + λx + λx + λx + λx are prime infinitely often for x,x,...,x, where x,x,...,x are natural numbers. Remark It is easy to see from Theorem . that primes pj in (.) are replaced by natural numbers xj and there exist infinitely many primes p,...,pr and p such that [μpk + ··· + μr+pk r+] = mpr, where m is a nonnegative integer (see []). 3 Outline of the proof Throughout this paper, p denotes a prime number, and xj denotes a natural number. δ is a sufficiently small positive number, ε is an arbitrarily small positive number. Constants, both explicit and implicit, in Landau or Vinogradov symbols may depend on λ,λ,...,λ. We write e(x) = exp(πix). We take X to be the basic parameter, a large real integer. Since at least one of the ratios λi/λj ( ≤ i < j ≤ ) is irrational, without loss of generality, we may assume that λ/λ is irrational. For the other cases, the only difference is in the following intermediate region, and we may deal with the same method in Section . Since λ/λ is irrational, there are infinitely many pairs of integers q, a with |λ/λ – a/q| ≥ q–, (p,q) = , q > and a ̸= . We choose p to be large in terms of λ,λ,...,λ, and make the following definitions. Put τ = N–+δ,T = N , L = logN, Q = (|λ|– + |λ|–)N–δ, [N–δ] = p and P = Nδ, where N ≍ X. Let ν be a positive real number, we define e(αx)dx. (.) Wang and Mu Journal of Inequalities and Applications ( 2017) 2017:170 Page 3 of 9 From (.) we have J =: � +∞ –∞ � i= Fi(λiα)G(–α)e � – α � K (α)dα ≤ logN � |λx +λx +λx +λx +λx +···+λx –p– |< ≤x,x≤X/,≤x,x≤X/,≤x,...,x≤X/,p≤N , which gives that
EARLIER¶9-13
variable x. We take X to be the basic parameter, a large real integer. Since at least one of the ratios λi/λj ( ≤ i < j ≤ ) is irrational, without loss of generality we may assume that λ/λ is irrational. For the other cases, the only difference is in the following intermediate region, and we may deal with the same method in Section . Since λ/λ is irrational, then there are infinitely many pairs of integers q, a with |λ/λ – a/q| ≤ q–, (a,q) = , q > and a ̸= . We choose q to be large in terms of λ,λ,...,λ and make the following definitions. N ≍ X, L = logN, � N–δ� = q, τ = N–+δ, Q = � |λ|– + |λ|–� N–δ, P = Nδ, T = N . Let ν be a positive real number, we define Kν(α) = ν �sinπνα πνα � , α ̸= , Kν() = ν, (.) Fi(α) = � ≤x≤X e � αx� , i = ,, Fj(α) = � ≤x≤X e � αx� , j = ,, Fk(α) = � ≤x≤X
LATER¶10-14
Theorem . Let λ,λ,...,λ be nonnegative real numbers, at least one of the ratios λi/λj ( ≤ i < j ≤ ) is rational. Then the integer parts of λx + λx + λx + λx + λx + λx + λx + λx are prime infinitely often for x,x,...,x, where x,x,...,x are natural numbers. Remark It is easy to see from Theorem . that primes pj in (.) are replaced by natural numbers xj and there exist infinitely many primes p,...,pr and p such that [μpk + ··· + μr+pk r+] = mpr, where m is a nonnegative integer (see []). 3 Outline of the proof Throughout this paper, p denotes a prime number, and xj denotes a natural number. δ is a sufficiently small positive number, ε is an arbitrarily small positive number. Constants, both explicit and implicit, in Landau or Vinogradov symbols may depend on λ,λ,...,λ. We write e(x) = exp(πix). We take X to be the basic parameter, a large real integer. Since at least one of the ratios λi/λj ( ≤ i < j ≤ ) is irrational, without loss of generality, we may assume that λ/λ is irrational. For the other cases, the only difference is in the following intermediate region, and we may deal with the same method in Section . Since λ/λ is irrational, there are infinitely many pairs of integers q, a with |λ/λ – a/q| ≥ q–, (p,q) = , q > and a ̸= . We choose p to be large in terms of λ,λ,...,λ, and make the following definitions. Put τ = N–+δ,T = N , L = logN, Q = (|λ|– + |λ|–)N–δ, [N–δ] = p and P = Nδ, where N ≍ X. Let ν be a positive real number, we define e(αx)dx. (.) Wang and Mu Journal of Inequalities and Applications ( 2017) 2017:170 Page 3 of 9 From (.) we have J =: � +∞ –∞ � i= Fi(λiα)G(–α)e � – α � K (α)dα ≤ logN � |λx +λx +λx +λx +λx +···+λx –p– |< ≤x,x≤X/,≤x,x≤X/,≤x,...,x≤X/,p≤N , which gives that
EARLIER¶26-30
Lemma . We have � |α|>N–+δ ����� � i= fi(λiα)g(–α) �����K (α)dα ≪ X – δ. Proof It follows from Vaughan [] that for α ̸= , fi(λiα) ≪ |α|– , i = ,, fj(λjα) ≪ |α|– , j = ,, fk(λkα) ≪ |α|– , k = ,...,, g(–α) ≪ |α|–. Thus � |α|>N–+δ ����� � i= fi(λiα)g(–α) �����K (α)dα ≪ � |α|>N–+δ |α|– dα ≪ X – δ. □ Lemma . We have � +∞ –∞ � i= fi(λiα)g(–α)e � – α � K (α)dα ≫ X . Proof From (.) one has x – ···x – · max � , – ����� � i= λixi – x – ����� � dxdx ··· dx. Lai Journal of Inequalities and Applications ( 2015) 2015:357 Page 6 of 9 Let |� i= λixi – x – | ≤ , then � i= λixi – ≤ x ≤ � i= λixi – . Based on � i= λixi – > , � i= λixi – < N, one may take
LATER¶11-16
are prime infinitely often for x,x,...,x, where x,x,...,x are natural numbers. Remark It is easy to see from Theorem . that primes pj in (.) are replaced by natural numbers xj and there exist infinitely many primes p,...,pr and p such that [μpk + ··· + μr+pk r+] = mpr, where m is a nonnegative integer (see []). 3 Outline of the proof Throughout this paper, p denotes a prime number, and xj denotes a natural number. δ is a sufficiently small positive number, ε is an arbitrarily small positive number. Constants, both explicit and implicit, in Landau or Vinogradov symbols may depend on λ,λ,...,λ. We write e(x) = exp(πix). We take X to be the basic parameter, a large real integer. Since at least one of the ratios λi/λj ( ≤ i < j ≤ ) is irrational, without loss of generality, we may assume that λ/λ is irrational. For the other cases, the only difference is in the following intermediate region, and we may deal with the same method in Section . Since λ/λ is irrational, there are infinitely many pairs of integers q, a with |λ/λ – a/q| ≥ q–, (p,q) = , q > and a ̸= . We choose p to be large in terms of λ,λ,...,λ, and make the following definitions. Put τ = N–+δ,T = N , L = logN, Q = (|λ|– + |λ|–)N–δ, [N–δ] = p and P = Nδ, where N ≍ X. Let ν be a positive real number, we define e(αx)dx. (.) Wang and Mu Journal of Inequalities and Applications ( 2017) 2017:170 Page 3 of 9 From (.) we have J =: � +∞ –∞ � i= Fi(λiα)G(–α)e � – α � K (α)dα ≤ logN � |λx +λx +λx +λx +λx +···+λx –p– |< ≤x,x≤X/,≤x,x≤X/,≤x,...,x≤X/,p≤N , which gives that N (X) ≥ (logN)–J. Next comes the time to estimate J. As usual, we split the range of infinite integration into three sections, C = {α ∈ R : |α| ≤ τ}, D = {α ∈ R : τ < |α| ≤ P}, c = {α ∈ R : |α| > P} named the neighborhood of the origin, the intermediate region, and the trivial region, respectively. In Sections , and , we shall establish that J(C) ≫ X , J(D) = o(X ), and J(c) = o(X ). Thus J ≫ X , N (X) ≫ X L–, namely, under the conditions of Theorem ., ����λx + λx + λx + λx + λx + λx + λx + λx – p –
EARLIER¶29-33
x – ···x – · max � , – ����� � i= λixi – x – ����� � dxdx ··· dx. Lai Journal of Inequalities and Applications ( 2015) 2015:357 Page 6 of 9 Let |� i= λixi – x – | ≤ , then � i= λixi – ≤ x ≤ � i= λixi – . Based on � i= λixi – > , � i= λixi – < N, one may take λjX � � i= λi �– ≤ xj ≤ λjX � � i= λi �– , j = ,...,, hence � +∞ –∞ � i= fi(λiα)g(–α)e � – α � K (α)dα ≥ ,, � j= λj � � i= λi �– X . This completes the proof of Lemma .. □ 4 The intermediate region Lemma . We have � +∞ –∞ ��Fi(λiα) ��K (α)dα ≪ X + ε, i = ,, (.) � +∞ –∞ ��Fj(λjα) ��K (α)dα ≪ X+ ε, j = ,, (.) � +∞ –∞ ��Fk(λkα) ��K (α)dα ≪ X + ε, k = ,...,, (.)
LATER¶11-16
are prime infinitely often for x,x,...,x, where x,x,...,x are natural numbers. Remark It is easy to see from Theorem . that primes pj in (.) are replaced by natural numbers xj and there exist infinitely many primes p,...,pr and p such that [μpk + ··· + μr+pk r+] = mpr, where m is a nonnegative integer (see []). 3 Outline of the proof Throughout this paper, p denotes a prime number, and xj denotes a natural number. δ is a sufficiently small positive number, ε is an arbitrarily small positive number. Constants, both explicit and implicit, in Landau or Vinogradov symbols may depend on λ,λ,...,λ. We write e(x) = exp(πix). We take X to be the basic parameter, a large real integer. Since at least one of the ratios λi/λj ( ≤ i < j ≤ ) is irrational, without loss of generality, we may assume that λ/λ is irrational. For the other cases, the only difference is in the following intermediate region, and we may deal with the same method in Section . Since λ/λ is irrational, there are infinitely many pairs of integers q, a with |λ/λ – a/q| ≥ q–, (p,q) = , q > and a ̸= . We choose p to be large in terms of λ,λ,...,λ, and make the following definitions. Put τ = N–+δ,T = N , L = logN, Q = (|λ|– + |λ|–)N–δ, [N–δ] = p and P = Nδ, where N ≍ X. Let ν be a positive real number, we define e(αx)dx. (.) Wang and Mu Journal of Inequalities and Applications ( 2017) 2017:170 Page 3 of 9 From (.) we have J =: � +∞ –∞ � i= Fi(λiα)G(–α)e � – α � K (α)dα ≤ logN � |λx +λx +λx +λx +λx +···+λx –p– |< ≤x,x≤X/,≤x,x≤X/,≤x,...,x≤X/,p≤N , which gives that N (X) ≥ (logN)–J. Next comes the time to estimate J. As usual, we split the range of infinite integration into three sections, C = {α ∈ R : |α| ≤ τ}, D = {α ∈ R : τ < |α| ≤ P}, c = {α ∈ R : |α| > P} named the neighborhood of the origin, the intermediate region, and the trivial region, respectively. In Sections , and , we shall establish that J(C) ≫ X , J(D) = o(X ), and J(c) = o(X ). Thus J ≫ X , N (X) ≫ X L–, namely, under the conditions of Theorem ., ����λx + λx + λx + λx + λx + λx + λx + λx – p –
EARLIER¶38-42
����aq λ λ – aq ���� ≤ ���� a/q λα qq � λα – a q ����� + ���� a/q λα qq � λα – a q ����� ≪ PQ– < q. We recall that q was chosen as the denominator of a convergent to the continued fraction for λ/λ. Thus, by Legendre’s law of best approximation, we have |q′ λ λ – a′| > q for all integers a′, q′ with ≤ q′ < q, thus |aq| ≥ q = [N–δ]. However, from (.) we have |aq| ≪ qqP ≪ Nδ, this is a contradiction. We have thus established that for at least one i, P < qi ≪ Q. Hence Lemma . gives the desired inequality for W(α). □ Lemma . We have � D � i= Fi(λiα)G(–α)e � – α � K (α)dα ≪ X – δ+ε. Proof By Lemmas ., . and Hölder’s inequality, we have � D � � X + ε� (NL) ≪ X – δ+ε. □ 5 The trivial region Lemma . (Lemma of []) Let V(α) = �e(αf (x,...,xm)), where f is any real function and the summation is over any finite set of values of x,...,xm. Then, for any A > , we have � |α|>A ��V(α) ��Kν(α)dα ≤ A � ∞ –∞ ��V(α) ��Kν(α)dα. Lemma . We have � c � i= Fi(λiα)G(–α)e � – α � K (α)dα ≪ X –δ+ε. Proof By Lemmas ., . and Schwarz’s inequality, we have
LATER¶11-16
are prime infinitely often for x,x,...,x, where x,x,...,x are natural numbers. Remark It is easy to see from Theorem . that primes pj in (.) are replaced by natural numbers xj and there exist infinitely many primes p,...,pr and p such that [μpk + ··· + μr+pk r+] = mpr, where m is a nonnegative integer (see []). 3 Outline of the proof Throughout this paper, p denotes a prime number, and xj denotes a natural number. δ is a sufficiently small positive number, ε is an arbitrarily small positive number. Constants, both explicit and implicit, in Landau or Vinogradov symbols may depend on λ,λ,...,λ. We write e(x) = exp(πix). We take X to be the basic parameter, a large real integer. Since at least one of the ratios λi/λj ( ≤ i < j ≤ ) is irrational, without loss of generality, we may assume that λ/λ is irrational. For the other cases, the only difference is in the following intermediate region, and we may deal with the same method in Section . Since λ/λ is irrational, there are infinitely many pairs of integers q, a with |λ/λ – a/q| ≥ q–, (p,q) = , q > and a ̸= . We choose p to be large in terms of λ,λ,...,λ, and make the following definitions. Put τ = N–+δ,T = N , L = logN, Q = (|λ|– + |λ|–)N–δ, [N–δ] = p and P = Nδ, where N ≍ X. Let ν be a positive real number, we define e(αx)dx. (.) Wang and Mu Journal of Inequalities and Applications ( 2017) 2017:170 Page 3 of 9 From (.) we have J =: � +∞ –∞ � i= Fi(λiα)G(–α)e � – α � K (α)dα ≤ logN � |λx +λx +λx +λx +λx +···+λx –p– |< ≤x,x≤X/,≤x,x≤X/,≤x,...,x≤X/,p≤N , which gives that N (X) ≥ (logN)–J. Next comes the time to estimate J. As usual, we split the range of infinite integration into three sections, C = {α ∈ R : |α| ≤ τ}, D = {α ∈ R : τ < |α| ≤ P}, c = {α ∈ R : |α| > P} named the neighborhood of the origin, the intermediate region, and the trivial region, respectively. In Sections , and , we shall establish that J(C) ≫ X , J(D) = o(X ), and J(c) = o(X ). Thus J ≫ X , N (X) ≫ X L–, namely, under the conditions of Theorem ., ����λx + λx + λx + λx + λx + λx + λx + λx – p –
EARLIER¶40-44
Lemma . We have � D � i= Fi(λiα)G(–α)e � – α � K (α)dα ≪ X – δ+ε. Proof By Lemmas ., . and Hölder’s inequality, we have � D � � X + ε� (NL) ≪ X – δ+ε. □ 5 The trivial region Lemma . (Lemma of []) Let V(α) = �e(αf (x,...,xm)), where f is any real function and the summation is over any finite set of values of x,...,xm. Then, for any A > , we have � |α|>A ��V(α) ��Kν(α)dα ≤ A � ∞ –∞ ��V(α) ��Kν(α)dα. Lemma . We have � c � i= Fi(λiα)G(–α)e � – α � K (α)dα ≪ X –δ+ε. Proof By Lemmas ., . and Schwarz’s inequality, we have (NL) ≪ X –δ+ε. □ 6 The proof of Theorem 1.1 From Lemmas ., . and . we conclude that J(C) ≫ X . From Lemma . it follows that J(D) = o(X ). From Lemma . we have J(c) = o(X ). Thus J ≫ X , N (X) ≫ X L–, namely, under conditions of Theorem ., ����λx + λx + λx + λx + λx + ··· + λx – p – ���� <
LATER¶11-16
are prime infinitely often for x,x,...,x, where x,x,...,x are natural numbers. Remark It is easy to see from Theorem . that primes pj in (.) are replaced by natural numbers xj and there exist infinitely many primes p,...,pr and p such that [μpk + ··· + μr+pk r+] = mpr, where m is a nonnegative integer (see []). 3 Outline of the proof Throughout this paper, p denotes a prime number, and xj denotes a natural number. δ is a sufficiently small positive number, ε is an arbitrarily small positive number. Constants, both explicit and implicit, in Landau or Vinogradov symbols may depend on λ,λ,...,λ. We write e(x) = exp(πix). We take X to be the basic parameter, a large real integer. Since at least one of the ratios λi/λj ( ≤ i < j ≤ ) is irrational, without loss of generality, we may assume that λ/λ is irrational. For the other cases, the only difference is in the following intermediate region, and we may deal with the same method in Section . Since λ/λ is irrational, there are infinitely many pairs of integers q, a with |λ/λ – a/q| ≥ q–, (p,q) = , q > and a ̸= . We choose p to be large in terms of λ,λ,...,λ, and make the following definitions. Put τ = N–+δ,T = N , L = logN, Q = (|λ|– + |λ|–)N–δ, [N–δ] = p and P = Nδ, where N ≍ X. Let ν be a positive real number, we define e(αx)dx. (.) Wang and Mu Journal of Inequalities and Applications ( 2017) 2017:170 Page 3 of 9 From (.) we have J =: � +∞ –∞ � i= Fi(λiα)G(–α)e � – α � K (α)dα ≤ logN � |λx +λx +λx +λx +λx +···+λx –p– |< ≤x,x≤X/,≤x,x≤X/,≤x,...,x≤X/,p≤N , which gives that N (X) ≥ (logN)–J. Next comes the time to estimate J. As usual, we split the range of infinite integration into three sections, C = {α ∈ R : |α| ≤ τ}, D = {α ∈ R : τ < |α| ≤ P}, c = {α ∈ R : |α| > P} named the neighborhood of the origin, the intermediate region, and the trivial region, respectively. In Sections , and , we shall establish that J(C) ≫ X , J(D) = o(X ), and J(c) = o(X ). Thus J ≫ X , N (X) ≫ X L–, namely, under the conditions of Theorem ., ����λx + λx + λx + λx + λx + λx + λx + λx – p –
EARLIER¶42-46
� |α|>A ��V(α) ��Kν(α)dα ≤ A � ∞ –∞ ��V(α) ��Kν(α)dα. Lemma . We have � c � i= Fi(λiα)G(–α)e � – α � K (α)dα ≪ X –δ+ε. Proof By Lemmas ., . and Schwarz’s inequality, we have (NL) ≪ X –δ+ε. □ 6 The proof of Theorem 1.1 From Lemmas ., . and . we conclude that J(C) ≫ X . From Lemma . it follows that J(D) = o(X ). From Lemma . we have J(c) = o(X ). Thus J ≫ X , N (X) ≫ X L–, namely, under conditions of Theorem ., ����λx + λx + λx + λx + λx + ··· + λx – p – ���� < (.) Lai Journal of Inequalities and Applications ( 2015) 2015:357 Page 9 of 9 has infinitely many solutions in positive integers x,x,...,x and prime p. It is evident from (.) that p < λx + λx + λx + λx + λx + ··· + λx < p + , and hence � λx + λx + λx + λx + λx + ··· + λx � = p. The proof of Theorem . is complete.
LATER¶14-18
α � K (α)dα ≤ logN � |λx +λx +λx +λx +λx +···+λx –p– |< ≤x,x≤X/,≤x,x≤X/,≤x,...,x≤X/,p≤N , which gives that N (X) ≥ (logN)–J. Next comes the time to estimate J. As usual, we split the range of infinite integration into three sections, C = {α ∈ R : |α| ≤ τ}, D = {α ∈ R : τ < |α| ≤ P}, c = {α ∈ R : |α| > P} named the neighborhood of the origin, the intermediate region, and the trivial region, respectively. In Sections , and , we shall establish that J(C) ≫ X , J(D) = o(X ), and J(c) = o(X ). Thus J ≫ X , N (X) ≫ X L–, namely, under the conditions of Theorem ., ����λx + λx + λx + λx + λx + λx + λx + λx – p – ���� < (.) has infinitely many solutions in positive integers x,x,...,x and prime p. From (.) we have p < λx + λx + λx + λx + λx + λx + λx + λx < p + , which gives that � λx + λx + λx + λx + λx + λx + λx + λx � = p. The proof of Theorem . is complete. 4 The neighborhood of the origin Lemma . (see [], Theorem .) Let (a,q) = . If α = a/q + β, then we have
EARLIER¶15-19
g(α) = � N e(αx)dx. It follows from (.) that Kν(α) ≪ min � ν,ν–|α|–� , (.) � +∞ –∞ e(αy)Kν(α)dα = max � , – ν–|y| � . (.) From (.) it is clear that |< ≤x,x≤X/,≤x,x≤X/,≤x,...,x≤X/,p≤N =: (logN)N (X), thus N (X) ≥ (logN)–J. To estimate J, we split the range of infinite integration into three sections, traditional named the neighborhood of the origin C = {α ∈ R : |α| ≤ τ}, the intermediate region D = {α ∈ R : τ < |α| ≤ P} and the trivial region c = {α ∈ R : |α| > P}. 3 The neighborhood of the origin Lemma . If α = a/q + β, where (a,q) = , then � ≤x≤N/t e � αxt� = q– q � m= e � amt/q �� N/t e � βyt� dy + O � q/+ε� + N|β| �� . Proof This is Theorem . of []. □ If |α| ∈ C, by Lemma ., taking a = , q = , then Fi(α) = fi(α) + O � Xδ� , i = ,,...,. (.) Lemma . Let ρ = β + iγ be a typical zero of the Riemann zeta function, C be a positive constant, I(α) = � |γ |≤T,β≥ � n≤N nρ–e(nα), J(α) = O �� + |α|N � N LC� , then G(α) = g(α) – I(α) + J(α), (.) � – ��I(α) �� dα ≪ N exp � –L � , (.)
LATER¶17-21
���� < (.) has infinitely many solutions in positive integers x,x,...,x and prime p. From (.) we have p < λx + λx + λx + λx + λx + λx + λx + λx < p + , which gives that � λx + λx + λx + λx + λx + λx + λx + λx � = p. The proof of Theorem . is complete. 4 The neighborhood of the origin Lemma . (see [], Theorem .) Let (a,q) = . If α = a/q + β, then we have � ≤x≤N/t e � αxt� = q– q � m= e � amt/q �� N/t e � βyt� dy + O � q/+ε� + N|β| �� . Lemma . immediately gives that Fi(α) = fi(α) + O � Xδ� , (.) where |α| ∈ C and i = ,,...,. Wang and Mu Journal of Inequalities and Applications ( 2017) 2017:170 Page 4 of 9 Lemma . (see [], Lemma and Remark ) Let I(α) = � |γ |≤T,β≥ � n≤N nρ–e(nα), J(α) = O �� + |α|N � N LC� , where C is a positive constant and ρ = β +iγ is a typical zero of the Riemann zeta function. Then we have
EARLIER¶17-21
3 The neighborhood of the origin Lemma . If α = a/q + β, where (a,q) = , then � ≤x≤N/t e � αxt� = q– q � m= e � amt/q �� N/t e � βyt� dy + O � q/+ε� + N|β| �� . Proof This is Theorem . of []. □ If |α| ∈ C, by Lemma ., taking a = , q = , then Fi(α) = fi(α) + O � Xδ� , i = ,,...,. (.) Lemma . Let ρ = β + iγ be a typical zero of the Riemann zeta function, C be a positive constant, I(α) = � |γ |≤T,β≥ � n≤N nρ–e(nα), J(α) = O �� + |α|N � N LC� , then G(α) = g(α) – I(α) + J(α), (.) � – ��I(α) �� dα ≪ N exp � –L � , (.) � τ –τ ��J(α) �� dα ≪ N exp � –L � . (.) Proof Equations (.), (.), (.) can be seen from Lemma , () and () given by Vaughan []. □ Lai Journal of Inequalities and Applications ( 2015) 2015:357 Page 4 of 9 Lemma . We have � – ��fi(α) �� dα ≪ X– , i = ,, � – ��fj(α) �� dα ≪ X– , j = ,, � – ��fk(α) �� dα ≪ X– , k = ,...,.
LATER¶19-23
� ≤x≤N/t e � αxt� = q– q � m= e � amt/q �� N/t e � βyt� dy + O � q/+ε� + N|β| �� . Lemma . immediately gives that Fi(α) = fi(α) + O � Xδ� , (.) where |α| ∈ C and i = ,,...,. Wang and Mu Journal of Inequalities and Applications ( 2017) 2017:170 Page 4 of 9 Lemma . (see [], Lemma and Remark ) Let I(α) = � |γ |≤T,β≥ � n≤N nρ–e(nα), J(α) = O �� + |α|N � N LC� , where C is a positive constant and ρ = β +iγ is a typical zero of the Riemann zeta function. Then we have � – ��I(α) �� dα ≪ N exp � –L � , � τ –τ ��J(α) �� dα ≪ N exp � –L � and G(α) = g(α) – I(α) + J(α). Lemma . (see [], Lemma ) For i = ,, j = ,,, k = ,,, we have � – ��fi(α) �� dα ≪ X– , � – ��fj(α) �� dα ≪ X– , � – ��fk(α) �� dα ≪ X– . Lemma . We have � C K (α) �����
EARLIER¶17-21
3 The neighborhood of the origin Lemma . If α = a/q + β, where (a,q) = , then � ≤x≤N/t e � αxt� = q– q � m= e � amt/q �� N/t e � βyt� dy + O � q/+ε� + N|β| �� . Proof This is Theorem . of []. □ If |α| ∈ C, by Lemma ., taking a = , q = , then Fi(α) = fi(α) + O � Xδ� , i = ,,...,. (.) Lemma . Let ρ = β + iγ be a typical zero of the Riemann zeta function, C be a positive constant, I(α) = � |γ |≤T,β≥ � n≤N nρ–e(nα), J(α) = O �� + |α|N � N LC� , then G(α) = g(α) – I(α) + J(α), (.) � – ��I(α) �� dα ≪ N exp � –L � , (.) � τ –τ ��J(α) �� dα ≪ N exp � –L � . (.) Proof Equations (.), (.), (.) can be seen from Lemma , () and () given by Vaughan []. □ Lai Journal of Inequalities and Applications ( 2015) 2015:357 Page 4 of 9 Lemma . We have � – ��fi(α) �� dα ≪ X– , i = ,, � – ��fj(α) �� dα ≪ X– , j = ,, � – ��fk(α) �� dα ≪ X– , k = ,...,.
LATER¶20-24
where |α| ∈ C and i = ,,...,. Wang and Mu Journal of Inequalities and Applications ( 2017) 2017:170 Page 4 of 9 Lemma . (see [], Lemma and Remark ) Let I(α) = � |γ |≤T,β≥ � n≤N nρ–e(nα), J(α) = O �� + |α|N � N LC� , where C is a positive constant and ρ = β +iγ is a typical zero of the Riemann zeta function. Then we have � – ��I(α) �� dα ≪ N exp � –L � , � τ –τ ��J(α) �� dα ≪ N exp � –L � and G(α) = g(α) – I(α) + J(α). Lemma . (see [], Lemma ) For i = ,, j = ,,, k = ,,, we have � – ��fi(α) �� dα ≪ X– , � – ��fj(α) �� dα ≪ X– , � – ��fk(α) �� dα ≪ X– . Lemma . We have � C K (α) ����� � i= Fi(λiα)G(–α) – � i= fi(λiα)g(–α) �����dα ≪ X L . Proof It is obvious that Fi(λiα) ≪ X , fi(λiα) ≪ X , Fj(λjα) ≪ X , fj(λjα) ≪ X ,
EARLIER¶18-22
Proof This is Theorem . of []. □ If |α| ∈ C, by Lemma ., taking a = , q = , then Fi(α) = fi(α) + O � Xδ� , i = ,,...,. (.) Lemma . Let ρ = β + iγ be a typical zero of the Riemann zeta function, C be a positive constant, I(α) = � |γ |≤T,β≥ � n≤N nρ–e(nα), J(α) = O �� + |α|N � N LC� , then G(α) = g(α) – I(α) + J(α), (.) � – ��I(α) �� dα ≪ N exp � –L � , (.) � τ –τ ��J(α) �� dα ≪ N exp � –L � . (.) Proof Equations (.), (.), (.) can be seen from Lemma , () and () given by Vaughan []. □ Lai Journal of Inequalities and Applications ( 2015) 2015:357 Page 4 of 9 Lemma . We have � – ��fi(α) �� dα ≪ X– , i = ,, � – ��fj(α) �� dα ≪ X– , j = ,, � – ��fk(α) �� dα ≪ X– , k = ,...,. Proof These results are from Lemma of []. □ Lemma . We have � C ����� � i= Fi(λiα)G(–α) – � i= fi(λiα)g(–α) �����K (α)dα ≪ X L–. Proof It is obvious that
LATER¶20-24
where |α| ∈ C and i = ,,...,. Wang and Mu Journal of Inequalities and Applications ( 2017) 2017:170 Page 4 of 9 Lemma . (see [], Lemma and Remark ) Let I(α) = � |γ |≤T,β≥ � n≤N nρ–e(nα), J(α) = O �� + |α|N � N LC� , where C is a positive constant and ρ = β +iγ is a typical zero of the Riemann zeta function. Then we have � – ��I(α) �� dα ≪ N exp � –L � , � τ –τ ��J(α) �� dα ≪ N exp � –L � and G(α) = g(α) – I(α) + J(α). Lemma . (see [], Lemma ) For i = ,, j = ,,, k = ,,, we have � – ��fi(α) �� dα ≪ X– , � – ��fj(α) �� dα ≪ X– , � – ��fk(α) �� dα ≪ X– . Lemma . We have � C K (α) ����� � i= Fi(λiα)G(–α) – � i= fi(λiα)g(–α) �����dα ≪ X L . Proof It is obvious that Fi(λiα) ≪ X , fi(λiα) ≪ X , Fj(λjα) ≪ X , fj(λjα) ≪ X ,
EARLIER¶19-23
I(α) = � |γ |≤T,β≥ � n≤N nρ–e(nα), J(α) = O �� + |α|N � N LC� , then G(α) = g(α) – I(α) + J(α), (.) � – ��I(α) �� dα ≪ N exp � –L � , (.) � τ –τ ��J(α) �� dα ≪ N exp � –L � . (.) Proof Equations (.), (.), (.) can be seen from Lemma , () and () given by Vaughan []. □ Lai Journal of Inequalities and Applications ( 2015) 2015:357 Page 4 of 9 Lemma . We have � – ��fi(α) �� dα ≪ X– , i = ,, � – ��fj(α) �� dα ≪ X– , j = ,, � – ��fk(α) �� dα ≪ X– , k = ,...,. Proof These results are from Lemma of []. □ Lemma . We have � C ����� � i= Fi(λiα)G(–α) – � i= fi(λiα)g(–α) �����K (α)dα ≪ X L–. Proof It is obvious that Fi(λiα) ≪ X , fi(λiα) ≪ X , i = ,, Fj(λjα) ≪ X , fj(λjα) ≪ X , j = ,, Fk(λkα) ≪ X , fk(λkα) ≪ X , k = ,...,, G(–α) ≪ N, g(–α) ≪ N,
LATER¶20-25
where |α| ∈ C and i = ,,...,. Wang and Mu Journal of Inequalities and Applications ( 2017) 2017:170 Page 4 of 9 Lemma . (see [], Lemma and Remark ) Let I(α) = � |γ |≤T,β≥ � n≤N nρ–e(nα), J(α) = O �� + |α|N � N LC� , where C is a positive constant and ρ = β +iγ is a typical zero of the Riemann zeta function. Then we have � – ��I(α) �� dα ≪ N exp � –L � , � τ –τ ��J(α) �� dα ≪ N exp � –L � and G(α) = g(α) – I(α) + J(α). Lemma . (see [], Lemma ) For i = ,, j = ,,, k = ,,, we have � – ��fi(α) �� dα ≪ X– , � – ��fj(α) �� dα ≪ X– , � – ��fk(α) �� dα ≪ X– . Lemma . We have � C K (α) ����� � i= Fi(λiα)G(–α) – � i= fi(λiα)g(–α) �����dα ≪ X L . Proof It is obvious that Fi(λiα) ≪ X , fi(λiα) ≪ X , Fj(λjα) ≪ X , fj(λjα) ≪ X , Fk(λkα) ≪ X , fk(λkα) ≪ X , G(–α) ≪ N and g(–α) ≪ N hold for i = ,, j = ,, and k = ,,. By (.), Lemmas . and ., we have � C ����� � F(λα) – f(λα) � �
EARLIER¶19-23
I(α) = � |γ |≤T,β≥ � n≤N nρ–e(nα), J(α) = O �� + |α|N � N LC� , then G(α) = g(α) – I(α) + J(α), (.) � – ��I(α) �� dα ≪ N exp � –L � , (.) � τ –τ ��J(α) �� dα ≪ N exp � –L � . (.) Proof Equations (.), (.), (.) can be seen from Lemma , () and () given by Vaughan []. □ Lai Journal of Inequalities and Applications ( 2015) 2015:357 Page 4 of 9 Lemma . We have � – ��fi(α) �� dα ≪ X– , i = ,, � – ��fj(α) �� dα ≪ X– , j = ,, � – ��fk(α) �� dα ≪ X– , k = ,...,. Proof These results are from Lemma of []. □ Lemma . We have � C ����� � i= Fi(λiα)G(–α) – � i= fi(λiα)g(–α) �����K (α)dα ≪ X L–. Proof It is obvious that Fi(λiα) ≪ X , fi(λiα) ≪ X , i = ,, Fj(λjα) ≪ X , fj(λjα) ≪ X , j = ,, Fk(λkα) ≪ X , fk(λkα) ≪ X , k = ,...,, G(–α) ≪ N, g(–α) ≪ N,
LATER¶21-25
I(α) = � |γ |≤T,β≥ � n≤N nρ–e(nα), J(α) = O �� + |α|N � N LC� , where C is a positive constant and ρ = β +iγ is a typical zero of the Riemann zeta function. Then we have � – ��I(α) �� dα ≪ N exp � –L � , � τ –τ ��J(α) �� dα ≪ N exp � –L � and G(α) = g(α) – I(α) + J(α). Lemma . (see [], Lemma ) For i = ,, j = ,,, k = ,,, we have � – ��fi(α) �� dα ≪ X– , � – ��fj(α) �� dα ≪ X– , � – ��fk(α) �� dα ≪ X– . Lemma . We have � C K (α) ����� � i= Fi(λiα)G(–α) – � i= fi(λiα)g(–α) �����dα ≪ X L . Proof It is obvious that Fi(λiα) ≪ X , fi(λiα) ≪ X , Fj(λjα) ≪ X , fj(λjα) ≪ X , Fk(λkα) ≪ X , fk(λkα) ≪ X , G(–α) ≪ N and g(–α) ≪ N hold for i = ,, j = ,, and k = ,,. By (.), Lemmas . and ., we have � C ����� � F(λα) – f(λα) � �
EARLIER¶19-23
I(α) = � |γ |≤T,β≥ � n≤N nρ–e(nα), J(α) = O �� + |α|N � N LC� , then G(α) = g(α) – I(α) + J(α), (.) � – ��I(α) �� dα ≪ N exp � –L � , (.) � τ –τ ��J(α) �� dα ≪ N exp � –L � . (.) Proof Equations (.), (.), (.) can be seen from Lemma , () and () given by Vaughan []. □ Lai Journal of Inequalities and Applications ( 2015) 2015:357 Page 4 of 9 Lemma . We have � – ��fi(α) �� dα ≪ X– , i = ,, � – ��fj(α) �� dα ≪ X– , j = ,, � – ��fk(α) �� dα ≪ X– , k = ,...,. Proof These results are from Lemma of []. □ Lemma . We have � C ����� � i= Fi(λiα)G(–α) – � i= fi(λiα)g(–α) �����K (α)dα ≪ X L–. Proof It is obvious that Fi(λiα) ≪ X , fi(λiα) ≪ X , i = ,, Fj(λjα) ≪ X , fj(λjα) ≪ X , j = ,, Fk(λkα) ≪ X , fk(λkα) ≪ X , k = ,...,, G(–α) ≪ N, g(–α) ≪ N,
LATER¶21-25
I(α) = � |γ |≤T,β≥ � n≤N nρ–e(nα), J(α) = O �� + |α|N � N LC� , where C is a positive constant and ρ = β +iγ is a typical zero of the Riemann zeta function. Then we have � – ��I(α) �� dα ≪ N exp � –L � , � τ –τ ��J(α) �� dα ≪ N exp � –L � and G(α) = g(α) – I(α) + J(α). Lemma . (see [], Lemma ) For i = ,, j = ,,, k = ,,, we have � – ��fi(α) �� dα ≪ X– , � – ��fj(α) �� dα ≪ X– , � – ��fk(α) �� dα ≪ X– . Lemma . We have � C K (α) ����� � i= Fi(λiα)G(–α) – � i= fi(λiα)g(–α) �����dα ≪ X L . Proof It is obvious that Fi(λiα) ≪ X , fi(λiα) ≪ X , Fj(λjα) ≪ X , fj(λjα) ≪ X , Fk(λkα) ≪ X , fk(λkα) ≪ X , G(–α) ≪ N and g(–α) ≪ N hold for i = ,, j = ,, and k = ,,. By (.), Lemmas . and ., we have � C ����� � F(λα) – f(λα) � �
EARLIER¶20-25
� τ –τ ��J(α) �� dα ≪ N exp � –L � . (.) Proof Equations (.), (.), (.) can be seen from Lemma , () and () given by Vaughan []. □ Lai Journal of Inequalities and Applications ( 2015) 2015:357 Page 4 of 9 Lemma . We have � – ��fi(α) �� dα ≪ X– , i = ,, � – ��fj(α) �� dα ≪ X– , j = ,, � – ��fk(α) �� dα ≪ X– , k = ,...,. Proof These results are from Lemma of []. □ Lemma . We have � C ����� � i= Fi(λiα)G(–α) – � i= fi(λiα)g(–α) �����K (α)dα ≪ X L–. Proof It is obvious that Fi(λiα) ≪ X , fi(λiα) ≪ X , i = ,, Fj(λjα) ≪ X , fj(λjα) ≪ X , j = ,, Fk(λkα) ≪ X , fk(λkα) ≪ X , k = ,...,, G(–α) ≪ N, g(–α) ≪ N, i= fi(λiα)G(–α) + � i= fi(λiα) � G(–α) – g(–α) � . Then by (.), Lemmas . and ., we have � C ����� � F(λα) – f(λα) � � i= Fi(λiα)G(–α) �� C ��J(α) �� dα + � – ��I(α) �� dα � ≪ X X– � N exp � –L �� ≪ X L–. The other cases are similar, and the proof of Lemma . is completed. □
LATER¶22-26
� – ��I(α) �� dα ≪ N exp � –L � , � τ –τ ��J(α) �� dα ≪ N exp � –L � and G(α) = g(α) – I(α) + J(α). Lemma . (see [], Lemma ) For i = ,, j = ,,, k = ,,, we have � – ��fi(α) �� dα ≪ X– , � – ��fj(α) �� dα ≪ X– , � – ��fk(α) �� dα ≪ X– . Lemma . We have � C K (α) ����� � i= Fi(λiα)G(–α) – � i= fi(λiα)g(–α) �����dα ≪ X L . Proof It is obvious that Fi(λiα) ≪ X , fi(λiα) ≪ X , Fj(λjα) ≪ X , fj(λjα) ≪ X , Fk(λkα) ≪ X , fk(λkα) ≪ X , G(–α) ≪ N and g(–α) ≪ N hold for i = ,, j = ,, and k = ,,. By (.), Lemmas . and ., we have � C ����� � F(λα) – f(λα) � � ≪ X �� – ��f(λα) �� dα � �� C ��J(α) �� dα + � – ��I(α) �� dα � ≪ X L . The proofs of the other cases are similar, so we complete the proof of Lemma .. □
EARLIER¶21-25
Lemma . We have � – ��fi(α) �� dα ≪ X– , i = ,, � – ��fj(α) �� dα ≪ X– , j = ,, � – ��fk(α) �� dα ≪ X– , k = ,...,. Proof These results are from Lemma of []. □ Lemma . We have � C ����� � i= Fi(λiα)G(–α) – � i= fi(λiα)g(–α) �����K (α)dα ≪ X L–. Proof It is obvious that Fi(λiα) ≪ X , fi(λiα) ≪ X , i = ,, Fj(λjα) ≪ X , fj(λjα) ≪ X , j = ,, Fk(λkα) ≪ X , fk(λkα) ≪ X , k = ,...,, G(–α) ≪ N, g(–α) ≪ N, i= fi(λiα)G(–α) + � i= fi(λiα) � G(–α) – g(–α) � . Then by (.), Lemmas . and ., we have � C ����� � F(λα) – f(λα) � � i= Fi(λiα)G(–α) �� C ��J(α) �� dα + � – ��I(α) �� dα � ≪ X X– � N exp � –L �� ≪ X L–. The other cases are similar, and the proof of Lemma . is completed. □
LATER¶22-26
� – ��I(α) �� dα ≪ N exp � –L � , � τ –τ ��J(α) �� dα ≪ N exp � –L � and G(α) = g(α) – I(α) + J(α). Lemma . (see [], Lemma ) For i = ,, j = ,,, k = ,,, we have � – ��fi(α) �� dα ≪ X– , � – ��fj(α) �� dα ≪ X– , � – ��fk(α) �� dα ≪ X– . Lemma . We have � C K (α) ����� � i= Fi(λiα)G(–α) – � i= fi(λiα)g(–α) �����dα ≪ X L . Proof It is obvious that Fi(λiα) ≪ X , fi(λiα) ≪ X , Fj(λjα) ≪ X , fj(λjα) ≪ X , Fk(λkα) ≪ X , fk(λkα) ≪ X , G(–α) ≪ N and g(–α) ≪ N hold for i = ,, j = ,, and k = ,,. By (.), Lemmas . and ., we have � C ����� � F(λα) – f(λα) � � ≪ X �� – ��f(λα) �� dα � �� C ��J(α) �� dα + � – ��I(α) �� dα � ≪ X L . The proofs of the other cases are similar, so we complete the proof of Lemma .. □
EARLIER¶23-27
Fi(λiα) ≪ X , fi(λiα) ≪ X , i = ,, Fj(λjα) ≪ X , fj(λjα) ≪ X , j = ,, Fk(λkα) ≪ X , fk(λkα) ≪ X , k = ,...,, G(–α) ≪ N, g(–α) ≪ N, i= fi(λiα)G(–α) + � i= fi(λiα) � G(–α) – g(–α) � . Then by (.), Lemmas . and ., we have � C ����� � F(λα) – f(λα) � � i= Fi(λiα)G(–α) �� C ��J(α) �� dα + � – ��I(α) �� dα � ≪ X X– � N exp � –L �� ≪ X L–. The other cases are similar, and the proof of Lemma . is completed. □ Lemma . We have � |α|>N–+δ ����� � i= fi(λiα)g(–α) �����K (α)dα ≪ X – δ. Proof It follows from Vaughan [] that for α ̸= , fi(λiα) ≪ |α|– , i = ,, fj(λjα) ≪ |α|– , j = ,, fk(λkα) ≪ |α|– , k = ,...,, g(–α) ≪ |α|–. Thus � |α|>N–+δ ����� � i= fi(λiα)g(–α) �����K (α)dα ≪ �
LATER¶24-28
� i= Fi(λiα)G(–α) – � i= fi(λiα)g(–α) �����dα ≪ X L . Proof It is obvious that Fi(λiα) ≪ X , fi(λiα) ≪ X , Fj(λjα) ≪ X , fj(λjα) ≪ X , Fk(λkα) ≪ X , fk(λkα) ≪ X , G(–α) ≪ N and g(–α) ≪ N hold for i = ,, j = ,, and k = ,,. By (.), Lemmas . and ., we have � C ����� � F(λα) – f(λα) � � ≪ X �� – ��f(λα) �� dα � �� C ��J(α) �� dα + � – ��I(α) �� dα � ≪ X L . The proofs of the other cases are similar, so we complete the proof of Lemma .. □ Lemma . The following inequality holds. � |α|> N–δ K (α) ����� � i= fi(λiα)g(–α) �����dα ≪ X – δ. Proof For α ̸= , i = ,, j = ,,, k = ,,, we know that fi(λiα) ≪ |α|– , fj(λjα) ≪ |α|– , fk(λkα) ≪ |α|– , g(–α) ≪ |α|–. Thus � |α|> N–δ ����� � i= fi(λiα)g(–α) �����K (α)dα ≪ � |α|>
EARLIER¶26-30
Lemma . We have � |α|>N–+δ ����� � i= fi(λiα)g(–α) �����K (α)dα ≪ X – δ. Proof It follows from Vaughan [] that for α ̸= , fi(λiα) ≪ |α|– , i = ,, fj(λjα) ≪ |α|– , j = ,, fk(λkα) ≪ |α|– , k = ,...,, g(–α) ≪ |α|–. Thus � |α|>N–+δ ����� � i= fi(λiα)g(–α) �����K (α)dα ≪ � |α|>N–+δ |α|– dα ≪ X – δ. □ Lemma . We have � +∞ –∞ � i= fi(λiα)g(–α)e � – α � K (α)dα ≫ X . Proof From (.) one has x – ···x – · max � , – ����� � i= λixi – x – ����� � dxdx ··· dx. Lai Journal of Inequalities and Applications ( 2015) 2015:357 Page 6 of 9 Let |� i= λixi – x – | ≤ , then � i= λixi – ≤ x ≤ � i= λixi – . Based on � i= λixi – > , � i= λixi – < N, one may take
LATER¶27-31
Lemma . The following inequality holds. � |α|> N–δ K (α) ����� � i= fi(λiα)g(–α) �����dα ≪ X – δ. Proof For α ̸= , i = ,, j = ,,, k = ,,, we know that fi(λiα) ≪ |α|– , fj(λjα) ≪ |α|– , fk(λkα) ≪ |α|– , g(–α) ≪ |α|–. Thus � |α|> N–δ ����� � i= fi(λiα)g(–α) �����K (α)dα ≪ � |α|> N–δ |α|– dα ≪ X – δ. □ Lemma . The following inequality holds. � +∞ –∞ � i= fi(λiα)g(–α)e � – α � K (α)dα ≫ X . – ����� � i= λixi – x – ����� � dxdx ··· dx from (.). Wang and Mu Journal of Inequalities and Applications ( 2017) 2017:170 Page 6 of 9 Let |� i= λixi – x – | ≤ . Then we have � i= λixi – ≤ x ≤ � i= λixi – . By using � i= λixi – > and � i= λixi – < N, we obtain that λjX �
EARLIER¶29-33
x – ···x – · max � , – ����� � i= λixi – x – ����� � dxdx ··· dx. Lai Journal of Inequalities and Applications ( 2015) 2015:357 Page 6 of 9 Let |� i= λixi – x – | ≤ , then � i= λixi – ≤ x ≤ � i= λixi – . Based on � i= λixi – > , � i= λixi – < N, one may take λjX � � i= λi �– ≤ xj ≤ λjX � � i= λi �– , j = ,...,, hence � +∞ –∞ � i= fi(λiα)g(–α)e � – α � K (α)dα ≥ ,, � j= λj � � i= λi �– X . This completes the proof of Lemma .. □ 4 The intermediate region Lemma . We have � +∞ –∞ ��Fi(λiα) ��K (α)dα ≪ X + ε, i = ,, (.) � +∞ –∞ ��Fj(λjα) ��K (α)dα ≪ X+ ε, j = ,, (.) � +∞ –∞ ��Fk(λkα) ��K (α)dα ≪ X + ε, k = ,...,, (.)
LATER¶27-31
Lemma . The following inequality holds. � |α|> N–δ K (α) ����� � i= fi(λiα)g(–α) �����dα ≪ X – δ. Proof For α ̸= , i = ,, j = ,,, k = ,,, we know that fi(λiα) ≪ |α|– , fj(λjα) ≪ |α|– , fk(λkα) ≪ |α|– , g(–α) ≪ |α|–. Thus � |α|> N–δ ����� � i= fi(λiα)g(–α) �����K (α)dα ≪ � |α|> N–δ |α|– dα ≪ X – δ. □ Lemma . The following inequality holds. � +∞ –∞ � i= fi(λiα)g(–α)e � – α � K (α)dα ≫ X . – ����� � i= λixi – x – ����� � dxdx ··· dx from (.). Wang and Mu Journal of Inequalities and Applications ( 2017) 2017:170 Page 6 of 9 Let |� i= λixi – x – | ≤ . Then we have � i= λixi – ≤ x ≤ � i= λixi – . By using � i= λixi – > and � i= λixi – < N, we obtain that λjX �
EARLIER¶38-42
����aq λ λ – aq ���� ≤ ���� a/q λα qq � λα – a q ����� + ���� a/q λα qq � λα – a q ����� ≪ PQ– < q. We recall that q was chosen as the denominator of a convergent to the continued fraction for λ/λ. Thus, by Legendre’s law of best approximation, we have |q′ λ λ – a′| > q for all integers a′, q′ with ≤ q′ < q, thus |aq| ≥ q = [N–δ]. However, from (.) we have |aq| ≪ qqP ≪ Nδ, this is a contradiction. We have thus established that for at least one i, P < qi ≪ Q. Hence Lemma . gives the desired inequality for W(α). □ Lemma . We have � D � i= Fi(λiα)G(–α)e � – α � K (α)dα ≪ X – δ+ε. Proof By Lemmas ., . and Hölder’s inequality, we have � D � � X + ε� (NL) ≪ X – δ+ε. □ 5 The trivial region Lemma . (Lemma of []) Let V(α) = �e(αf (x,...,xm)), where f is any real function and the summation is over any finite set of values of x,...,xm. Then, for any A > , we have � |α|>A ��V(α) ��Kν(α)dα ≤ A � ∞ –∞ ��V(α) ��Kν(α)dα. Lemma . We have � c � i= Fi(λiα)G(–α)e � – α � K (α)dα ≪ X –δ+ε. Proof By Lemmas ., . and Schwarz’s inequality, we have
LATER¶27-31
Lemma . The following inequality holds. � |α|> N–δ K (α) ����� � i= fi(λiα)g(–α) �����dα ≪ X – δ. Proof For α ̸= , i = ,, j = ,,, k = ,,, we know that fi(λiα) ≪ |α|– , fj(λjα) ≪ |α|– , fk(λkα) ≪ |α|– , g(–α) ≪ |α|–. Thus � |α|> N–δ ����� � i= fi(λiα)g(–α) �����K (α)dα ≪ � |α|> N–δ |α|– dα ≪ X – δ. □ Lemma . The following inequality holds. � +∞ –∞ � i= fi(λiα)g(–α)e � – α � K (α)dα ≫ X . – ����� � i= λixi – x – ����� � dxdx ··· dx from (.). Wang and Mu Journal of Inequalities and Applications ( 2017) 2017:170 Page 6 of 9 Let |� i= λixi – x – | ≤ . Then we have � i= λixi – ≤ x ≤ � i= λixi – . By using � i= λixi – > and � i= λixi – < N, we obtain that λjX �
EARLIER¶40-44
Lemma . We have � D � i= Fi(λiα)G(–α)e � – α � K (α)dα ≪ X – δ+ε. Proof By Lemmas ., . and Hölder’s inequality, we have � D � � X + ε� (NL) ≪ X – δ+ε. □ 5 The trivial region Lemma . (Lemma of []) Let V(α) = �e(αf (x,...,xm)), where f is any real function and the summation is over any finite set of values of x,...,xm. Then, for any A > , we have � |α|>A ��V(α) ��Kν(α)dα ≤ A � ∞ –∞ ��V(α) ��Kν(α)dα. Lemma . We have � c � i= Fi(λiα)G(–α)e � – α � K (α)dα ≪ X –δ+ε. Proof By Lemmas ., . and Schwarz’s inequality, we have (NL) ≪ X –δ+ε. □ 6 The proof of Theorem 1.1 From Lemmas ., . and . we conclude that J(C) ≫ X . From Lemma . it follows that J(D) = o(X ). From Lemma . we have J(c) = o(X ). Thus J ≫ X , N (X) ≫ X L–, namely, under conditions of Theorem ., ����λx + λx + λx + λx + λx + ··· + λx – p – ���� <
LATER¶27-31
Lemma . The following inequality holds. � |α|> N–δ K (α) ����� � i= fi(λiα)g(–α) �����dα ≪ X – δ. Proof For α ̸= , i = ,, j = ,,, k = ,,, we know that fi(λiα) ≪ |α|– , fj(λjα) ≪ |α|– , fk(λkα) ≪ |α|– , g(–α) ≪ |α|–. Thus � |α|> N–δ ����� � i= fi(λiα)g(–α) �����K (α)dα ≪ � |α|> N–δ |α|– dα ≪ X – δ. □ Lemma . The following inequality holds. � +∞ –∞ � i= fi(λiα)g(–α)e � – α � K (α)dα ≫ X . – ����� � i= λixi – x – ����� � dxdx ··· dx from (.). Wang and Mu Journal of Inequalities and Applications ( 2017) 2017:170 Page 6 of 9 Let |� i= λixi – x – | ≤ . Then we have � i= λixi – ≤ x ≤ � i= λixi – . By using � i= λixi – > and � i= λixi – < N, we obtain that λjX �
EARLIER¶29-33
x – ···x – · max � , – ����� � i= λixi – x – ����� � dxdx ··· dx. Lai Journal of Inequalities and Applications ( 2015) 2015:357 Page 6 of 9 Let |� i= λixi – x – | ≤ , then � i= λixi – ≤ x ≤ � i= λixi – . Based on � i= λixi – > , � i= λixi – < N, one may take λjX � � i= λi �– ≤ xj ≤ λjX � � i= λi �– , j = ,...,, hence � +∞ –∞ � i= fi(λiα)g(–α)e � – α � K (α)dα ≥ ,, � j= λj � � i= λi �– X . This completes the proof of Lemma .. □ 4 The intermediate region Lemma . We have � +∞ –∞ ��Fi(λiα) ��K (α)dα ≪ X + ε, i = ,, (.) � +∞ –∞ ��Fj(λjα) ��K (α)dα ≪ X+ ε, j = ,, (.) � +∞ –∞ ��Fk(λkα) ��K (α)dα ≪ X + ε, k = ,...,, (.)
LATER¶30-34
– ����� � i= λixi – x – ����� � dxdx ··· dx from (.). Wang and Mu Journal of Inequalities and Applications ( 2017) 2017:170 Page 6 of 9 Let |� i= λixi – x – | ≤ . Then we have � i= λixi – ≤ x ≤ � i= λixi – . By using � i= λixi – > and � i= λixi – < N, we obtain that λjX � � i= λi �– ≤ xj ≤ λjX � � i= λi �– , j = ,...,, and hence � +∞ –∞ � i= fi(λiα)g(–α)e � – α � K (α)dα ≥ , � j= λj � � i= λi �– X . Then we complete the proof of this lemma. □ 5 The intermediate region Lemma . We have � +∞ –∞ ��Fi(λiα) ��K (α)dα ≪ X + ε, � +∞ –∞ ��Fj(λjα) ��K (α)dα ≪ X+ ε, � +∞ –∞ ��Fk(λkα) ��K (α)dα ≪ X + ε and � +∞ –∞ ��G(–α) ��K (α)dα ≪ NL for i = ,, j = ,, and k = ,,.
EARLIER¶26-30
Lemma . We have � |α|>N–+δ ����� � i= fi(λiα)g(–α) �����K (α)dα ≪ X – δ. Proof It follows from Vaughan [] that for α ̸= , fi(λiα) ≪ |α|– , i = ,, fj(λjα) ≪ |α|– , j = ,, fk(λkα) ≪ |α|– , k = ,...,, g(–α) ≪ |α|–. Thus � |α|>N–+δ ����� � i= fi(λiα)g(–α) �����K (α)dα ≪ � |α|>N–+δ |α|– dα ≪ X – δ. □ Lemma . We have � +∞ –∞ � i= fi(λiα)g(–α)e � – α � K (α)dα ≫ X . Proof From (.) one has x – ···x – · max � , – ����� � i= λixi – x – ����� � dxdx ··· dx. Lai Journal of Inequalities and Applications ( 2015) 2015:357 Page 6 of 9 Let |� i= λixi – x – | ≤ , then � i= λixi – ≤ x ≤ � i= λixi – . Based on � i= λixi – > , � i= λixi – < N, one may take
LATER¶30-34
– ����� � i= λixi – x – ����� � dxdx ··· dx from (.). Wang and Mu Journal of Inequalities and Applications ( 2017) 2017:170 Page 6 of 9 Let |� i= λixi – x – | ≤ . Then we have � i= λixi – ≤ x ≤ � i= λixi – . By using � i= λixi – > and � i= λixi – < N, we obtain that λjX � � i= λi �– ≤ xj ≤ λjX � � i= λi �– , j = ,...,, and hence � +∞ –∞ � i= fi(λiα)g(–α)e � – α � K (α)dα ≥ , � j= λj � � i= λi �– X . Then we complete the proof of this lemma. □ 5 The intermediate region Lemma . We have � +∞ –∞ ��Fi(λiα) ��K (α)dα ≪ X + ε, � +∞ –∞ ��Fj(λjα) ��K (α)dα ≪ X+ ε, � +∞ –∞ ��Fk(λkα) ��K (α)dα ≪ X + ε and � +∞ –∞ ��G(–α) ��K (α)dα ≪ NL for i = ,, j = ,, and k = ,,.
EARLIER¶38-42
����aq λ λ – aq ���� ≤ ���� a/q λα qq � λα – a q ����� + ���� a/q λα qq � λα – a q ����� ≪ PQ– < q. We recall that q was chosen as the denominator of a convergent to the continued fraction for λ/λ. Thus, by Legendre’s law of best approximation, we have |q′ λ λ – a′| > q for all integers a′, q′ with ≤ q′ < q, thus |aq| ≥ q = [N–δ]. However, from (.) we have |aq| ≪ qqP ≪ Nδ, this is a contradiction. We have thus established that for at least one i, P < qi ≪ Q. Hence Lemma . gives the desired inequality for W(α). □ Lemma . We have � D � i= Fi(λiα)G(–α)e � – α � K (α)dα ≪ X – δ+ε. Proof By Lemmas ., . and Hölder’s inequality, we have � D � � X + ε� (NL) ≪ X – δ+ε. □ 5 The trivial region Lemma . (Lemma of []) Let V(α) = �e(αf (x,...,xm)), where f is any real function and the summation is over any finite set of values of x,...,xm. Then, for any A > , we have � |α|>A ��V(α) ��Kν(α)dα ≤ A � ∞ –∞ ��V(α) ��Kν(α)dα. Lemma . We have � c � i= Fi(λiα)G(–α)e � – α � K (α)dα ≪ X –δ+ε. Proof By Lemmas ., . and Schwarz’s inequality, we have
LATER¶30-34
– ����� � i= λixi – x – ����� � dxdx ··· dx from (.). Wang and Mu Journal of Inequalities and Applications ( 2017) 2017:170 Page 6 of 9 Let |� i= λixi – x – | ≤ . Then we have � i= λixi – ≤ x ≤ � i= λixi – . By using � i= λixi – > and � i= λixi – < N, we obtain that λjX � � i= λi �– ≤ xj ≤ λjX � � i= λi �– , j = ,...,, and hence � +∞ –∞ � i= fi(λiα)g(–α)e � – α � K (α)dα ≥ , � j= λj � � i= λi �– X . Then we complete the proof of this lemma. □ 5 The intermediate region Lemma . We have � +∞ –∞ ��Fi(λiα) ��K (α)dα ≪ X + ε, � +∞ –∞ ��Fj(λjα) ��K (α)dα ≪ X+ ε, � +∞ –∞ ��Fk(λkα) ��K (α)dα ≪ X + ε and � +∞ –∞ ��G(–α) ��K (α)dα ≪ NL for i = ,, j = ,, and k = ,,.
EARLIER¶40-44
Lemma . We have � D � i= Fi(λiα)G(–α)e � – α � K (α)dα ≪ X – δ+ε. Proof By Lemmas ., . and Hölder’s inequality, we have � D � � X + ε� (NL) ≪ X – δ+ε. □ 5 The trivial region Lemma . (Lemma of []) Let V(α) = �e(αf (x,...,xm)), where f is any real function and the summation is over any finite set of values of x,...,xm. Then, for any A > , we have � |α|>A ��V(α) ��Kν(α)dα ≤ A � ∞ –∞ ��V(α) ��Kν(α)dα. Lemma . We have � c � i= Fi(λiα)G(–α)e � – α � K (α)dα ≪ X –δ+ε. Proof By Lemmas ., . and Schwarz’s inequality, we have (NL) ≪ X –δ+ε. □ 6 The proof of Theorem 1.1 From Lemmas ., . and . we conclude that J(C) ≫ X . From Lemma . it follows that J(D) = o(X ). From Lemma . we have J(c) = o(X ). Thus J ≫ X , N (X) ≫ X L–, namely, under conditions of Theorem ., ����λx + λx + λx + λx + λx + ··· + λx – p – ���� <
LATER¶30-34
– ����� � i= λixi – x – ����� � dxdx ··· dx from (.). Wang and Mu Journal of Inequalities and Applications ( 2017) 2017:170 Page 6 of 9 Let |� i= λixi – x – | ≤ . Then we have � i= λixi – ≤ x ≤ � i= λixi – . By using � i= λixi – > and � i= λixi – < N, we obtain that λjX � � i= λi �– ≤ xj ≤ λjX � � i= λi �– , j = ,...,, and hence � +∞ –∞ � i= fi(λiα)g(–α)e � – α � K (α)dα ≥ , � j= λj � � i= λi �– X . Then we complete the proof of this lemma. □ 5 The intermediate region Lemma . We have � +∞ –∞ ��Fi(λiα) ��K (α)dα ≪ X + ε, � +∞ –∞ ��Fj(λjα) ��K (α)dα ≪ X+ ε, � +∞ –∞ ��Fk(λkα) ��K (α)dα ≪ X + ε and � +∞ –∞ ��G(–α) ��K (α)dα ≪ NL for i = ,, j = ,, and k = ,,.
EARLIER¶30-35
Let |� i= λixi – x – | ≤ , then � i= λixi – ≤ x ≤ � i= λixi – . Based on � i= λixi – > , � i= λixi – < N, one may take λjX � � i= λi �– ≤ xj ≤ λjX � � i= λi �– , j = ,...,, hence � +∞ –∞ � i= fi(λiα)g(–α)e � – α � K (α)dα ≥ ,, � j= λj � � i= λi �– X . This completes the proof of Lemma .. □ 4 The intermediate region Lemma . We have � +∞ –∞ ��Fi(λiα) ��K (α)dα ≪ X + ε, i = ,, (.) � +∞ –∞ ��Fj(λjα) ��K (α)dα ≪ X+ ε, j = ,, (.) � +∞ –∞ ��Fk(λkα) ��K (α)dα ≪ X + ε, k = ,...,, (.) � +∞ –∞ ��G(–α) ��K (α)dα ≪ NL. (.) Proof By (.) and Hua’s inequality, for i = ,, we have � +∞ –∞ ��Fi(λiα) ��K (α)dα ≪ +∞ � m=–∞ � m+ ε. The proofs of (.)-(.) are similar to (.). □ Lai Journal of Inequalities and Applications ( 2015) 2015:357 Page 7 of 9 Lemma . Suppose that (a,q) = , |α – a/q| ≤ q–, φ(x) = αxk + αxk– + ··· + αk–x + αk, then
LATER¶31-36
| ≤ . Then we have � i= λixi – ≤ x ≤ � i= λixi – . By using � i= λixi – > and � i= λixi – < N, we obtain that λjX � � i= λi �– ≤ xj ≤ λjX � � i= λi �– , j = ,...,, and hence � +∞ –∞ � i= fi(λiα)g(–α)e � – α � K (α)dα ≥ , � j= λj � � i= λi �– X . Then we complete the proof of this lemma. □ 5 The intermediate region Lemma . We have � +∞ –∞ ��Fi(λiα) ��K (α)dα ≪ X + ε, � +∞ –∞ ��Fj(λjα) ��K (α)dα ≪ X+ ε, � +∞ –∞ ��Fk(λkα) ��K (α)dα ≪ X + ε and � +∞ –∞ ��G(–α) ��K (α)dα ≪ NL for i = ,, j = ,, and k = ,,. ≪ � m= � m+ m ��Fj(λjα) �� dα + +∞ � m= m– � m+ m ��Fj(λjα) �� dα ≪ X+ ε from (.) and Hua’s inequality. The proofs of others are similar. So we omit them here. □ Lemma . (see [], Lemma . (Weyl’s inequality)) Suppose that ����α – a q ���� ≤ q , (a,q) = and φ(x) = αxk + αxk– + ··· + αk–x + αk. Then we have
EARLIER¶31-35
λjX � � i= λi �– ≤ xj ≤ λjX � � i= λi �– , j = ,...,, hence � +∞ –∞ � i= fi(λiα)g(–α)e � – α � K (α)dα ≥ ,, � j= λj � � i= λi �– X . This completes the proof of Lemma .. □ 4 The intermediate region Lemma . We have � +∞ –∞ ��Fi(λiα) ��K (α)dα ≪ X + ε, i = ,, (.) � +∞ –∞ ��Fj(λjα) ��K (α)dα ≪ X+ ε, j = ,, (.) � +∞ –∞ ��Fk(λkα) ��K (α)dα ≪ X + ε, k = ,...,, (.) � +∞ –∞ ��G(–α) ��K (α)dα ≪ NL. (.) Proof By (.) and Hua’s inequality, for i = ,, we have � +∞ –∞ ��Fi(λiα) ��K (α)dα ≪ +∞ � m=–∞ � m+ ε. The proofs of (.)-(.) are similar to (.). □ Lai Journal of Inequalities and Applications ( 2015) 2015:357 Page 7 of 9 Lemma . Suppose that (a,q) = , |α – a/q| ≤ q–, φ(x) = αxk + αxk– + ··· + αk–x + αk, then
LATER¶32-36
� i= λi �– ≤ xj ≤ λjX � � i= λi �– , j = ,...,, and hence � +∞ –∞ � i= fi(λiα)g(–α)e � – α � K (α)dα ≥ , � j= λj � � i= λi �– X . Then we complete the proof of this lemma. □ 5 The intermediate region Lemma . We have � +∞ –∞ ��Fi(λiα) ��K (α)dα ≪ X + ε, � +∞ –∞ ��Fj(λjα) ��K (α)dα ≪ X+ ε, � +∞ –∞ ��Fk(λkα) ��K (α)dα ≪ X + ε and � +∞ –∞ ��G(–α) ��K (α)dα ≪ NL for i = ,, j = ,, and k = ,,. ≪ � m= � m+ m ��Fj(λjα) �� dα + +∞ � m= m– � m+ m ��Fj(λjα) �� dα ≪ X+ ε from (.) and Hua’s inequality. The proofs of others are similar. So we omit them here. □ Lemma . (see [], Lemma . (Weyl’s inequality)) Suppose that ����α – a q ���� ≤ q , (a,q) = and φ(x) = αxk + αxk– + ··· + αk–x + αk. Then we have
EARLIER¶31-35
λjX � � i= λi �– ≤ xj ≤ λjX � � i= λi �– , j = ,...,, hence � +∞ –∞ � i= fi(λiα)g(–α)e � – α � K (α)dα ≥ ,, � j= λj � � i= λi �– X . This completes the proof of Lemma .. □ 4 The intermediate region Lemma . We have � +∞ –∞ ��Fi(λiα) ��K (α)dα ≪ X + ε, i = ,, (.) � +∞ –∞ ��Fj(λjα) ��K (α)dα ≪ X+ ε, j = ,, (.) � +∞ –∞ ��Fk(λkα) ��K (α)dα ≪ X + ε, k = ,...,, (.) � +∞ –∞ ��G(–α) ��K (α)dα ≪ NL. (.) Proof By (.) and Hua’s inequality, for i = ,, we have � +∞ –∞ ��Fi(λiα) ��K (α)dα ≪ +∞ � m=–∞ � m+ ε. The proofs of (.)-(.) are similar to (.). □ Lai Journal of Inequalities and Applications ( 2015) 2015:357 Page 7 of 9 Lemma . Suppose that (a,q) = , |α – a/q| ≤ q–, φ(x) = αxk + αxk– + ··· + αk–x + αk, then
LATER¶32-36
� i= λi �– ≤ xj ≤ λjX � � i= λi �– , j = ,...,, and hence � +∞ –∞ � i= fi(λiα)g(–α)e � – α � K (α)dα ≥ , � j= λj � � i= λi �– X . Then we complete the proof of this lemma. □ 5 The intermediate region Lemma . We have � +∞ –∞ ��Fi(λiα) ��K (α)dα ≪ X + ε, � +∞ –∞ ��Fj(λjα) ��K (α)dα ≪ X+ ε, � +∞ –∞ ��Fk(λkα) ��K (α)dα ≪ X + ε and � +∞ –∞ ��G(–α) ��K (α)dα ≪ NL for i = ,, j = ,, and k = ,,. ≪ � m= � m+ m ��Fj(λjα) �� dα + +∞ � m= m– � m+ m ��Fj(λjα) �� dα ≪ X+ ε from (.) and Hua’s inequality. The proofs of others are similar. So we omit them here. □ Lemma . (see [], Lemma . (Weyl’s inequality)) Suppose that ����α – a q ���� ≤ q , (a,q) = and φ(x) = αxk + αxk– + ··· + αk–x + αk. Then we have
EARLIER¶34-38
� +∞ –∞ ��G(–α) ��K (α)dα ≪ NL. (.) Proof By (.) and Hua’s inequality, for i = ,, we have � +∞ –∞ ��Fi(λiα) ��K (α)dα ≪ +∞ � m=–∞ � m+ ε. The proofs of (.)-(.) are similar to (.). □ Lai Journal of Inequalities and Applications ( 2015) 2015:357 Page 7 of 9 Lemma . Suppose that (a,q) = , |α – a/q| ≤ q–, φ(x) = αxk + αxk– + ··· + αk–x + αk, then M � x= e � φ(x) � ≪ M+ε� q– + M– + qM–k�–k . Proof This is Lemma . (Weyl’s inequality) of Vaughan []. □ Lemma . For every real number α ∈ D, let W(α) = min(|F(λα)|,|F(λα)|), then W(α) ≪ X – δ+ ε. Proof For α ∈ D and i = ,, we choose ai, qi such that |λiα – ai/qi| ≤ q– i Q– (.) with (ai,qi) = and ≤ qi ≤ Q. Firstly, we note that aa ̸= . Secondly, if q,q ≤ P, then ����aq λ λ – aq ���� ≤ ���� a/q λα qq � λα – a q ����� + ���� a/q λα qq � λα – a q ����� ≪ PQ– < q. We recall that q was chosen as the denominator of a convergent to the continued fraction for λ/λ. Thus, by Legendre’s law of best approximation, we have |q′ λ
LATER¶35-39
≪ � m= � m+ m ��Fj(λjα) �� dα + +∞ � m= m– � m+ m ��Fj(λjα) �� dα ≪ X+ ε from (.) and Hua’s inequality. The proofs of others are similar. So we omit them here. □ Lemma . (see [], Lemma . (Weyl’s inequality)) Suppose that ����α – a q ���� ≤ q , (a,q) = and φ(x) = αxk + αxk– + ··· + αk–x + αk. Then we have M � x= e � φ(x) � ≪ M+ε� q– + M– + qM–k�–k . Lemma . For every real number α ∈ D, we have W(α) ≪ X – δ+ ε, where W(α) = min ���G(τα) ��, ��G(τα) ��� . Proof For α ∈ D and i = ,, we choose ai, qi such that |λiα – ai/qi| ≤ q– i Q– with (ai,qi) = and ≤ qi ≤ Q. We note that aa ̸= . If q,q ≤ P, then ����aq λ λ – aq ���� ≤ ���� a/q λα qq � λα – a q ����� + ���� a/q λα qq � λα – a q ����� ≪ PQ– < q. We recall that q was chosen as the denominator of a convergent to the continued fraction for λ/λ. Thus, by Legendre’s law of best approximation, we have |q′ λ
EARLIER¶35-39
ε. The proofs of (.)-(.) are similar to (.). □ Lai Journal of Inequalities and Applications ( 2015) 2015:357 Page 7 of 9 Lemma . Suppose that (a,q) = , |α – a/q| ≤ q–, φ(x) = αxk + αxk– + ··· + αk–x + αk, then M � x= e � φ(x) � ≪ M+ε� q– + M– + qM–k�–k . Proof This is Lemma . (Weyl’s inequality) of Vaughan []. □ Lemma . For every real number α ∈ D, let W(α) = min(|F(λα)|,|F(λα)|), then W(α) ≪ X – δ+ ε. Proof For α ∈ D and i = ,, we choose ai, qi such that |λiα – ai/qi| ≤ q– i Q– (.) with (ai,qi) = and ≤ qi ≤ Q. Firstly, we note that aa ̸= . Secondly, if q,q ≤ P, then ����aq λ λ – aq ���� ≤ ���� a/q λα qq � λα – a q ����� + ���� a/q λα qq � λα – a q ����� ≪ PQ– < q. We recall that q was chosen as the denominator of a convergent to the continued fraction for λ/λ. Thus, by Legendre’s law of best approximation, we have |q′ λ λ – a′| > q for all integers a′, q′ with ≤ q′ < q, thus |aq| ≥ q = [N–δ]. However, from (.) we have |aq| ≪ qqP ≪ Nδ, this is a contradiction. We have thus established that for at least one i, P < qi ≪ Q. Hence Lemma . gives the desired inequality for W(α). □
LATER¶36-40
Lemma . (see [], Lemma . (Weyl’s inequality)) Suppose that ����α – a q ���� ≤ q , (a,q) = and φ(x) = αxk + αxk– + ··· + αk–x + αk. Then we have M � x= e � φ(x) � ≪ M+ε� q– + M– + qM–k�–k . Lemma . For every real number α ∈ D, we have W(α) ≪ X – δ+ ε, where W(α) = min ���G(τα) ��, ��G(τα) ��� . Proof For α ∈ D and i = ,, we choose ai, qi such that |λiα – ai/qi| ≤ q– i Q– with (ai,qi) = and ≤ qi ≤ Q. We note that aa ̸= . If q,q ≤ P, then ����aq λ λ – aq ���� ≤ ���� a/q λα qq � λα – a q ����� + ���� a/q λα qq � λα – a q ����� ≪ PQ– < q. We recall that q was chosen as the denominator of a convergent to the continued fraction for λ/λ. Thus, by Legendre’s law of best approximation, we have |q′ λ λ – a′| > q for all integers a′, q′ with ≤ q′ < q, thus |aq| ≥ q = [N–δ]. On the other hand, |aq| ≪ qqP ≪ Nδ, which is a contradiction. And so, for at least one i, P < qi ≪ Q. Hence, by Lemma ., we obtain the desired inequality for W(α). □
EARLIER¶35-41
ε. The proofs of (.)-(.) are similar to (.). □ Lai Journal of Inequalities and Applications ( 2015) 2015:357 Page 7 of 9 Lemma . Suppose that (a,q) = , |α – a/q| ≤ q–, φ(x) = αxk + αxk– + ··· + αk–x + αk, then M � x= e � φ(x) � ≪ M+ε� q– + M– + qM–k�–k . Proof This is Lemma . (Weyl’s inequality) of Vaughan []. □ Lemma . For every real number α ∈ D, let W(α) = min(|F(λα)|,|F(λα)|), then W(α) ≪ X – δ+ ε. Proof For α ∈ D and i = ,, we choose ai, qi such that |λiα – ai/qi| ≤ q– i Q– (.) with (ai,qi) = and ≤ qi ≤ Q. Firstly, we note that aa ̸= . Secondly, if q,q ≤ P, then ����aq λ λ – aq ���� ≤ ���� a/q λα qq � λα – a q ����� + ���� a/q λα qq � λα – a q ����� ≪ PQ– < q. We recall that q was chosen as the denominator of a convergent to the continued fraction for λ/λ. Thus, by Legendre’s law of best approximation, we have |q′ λ λ – a′| > q for all integers a′, q′ with ≤ q′ < q, thus |aq| ≥ q = [N–δ]. However, from (.) we have |aq| ≪ qqP ≪ Nδ, this is a contradiction. We have thus established that for at least one i, P < qi ≪ Q. Hence Lemma . gives the desired inequality for W(α). □ Lemma . We have � D � i= Fi(λiα)G(–α)e � – α � K (α)dα ≪ X – δ+ε. Proof By Lemmas ., . and Hölder’s inequality, we have � D � � X + ε� (NL) ≪ X – δ+ε. □ 5 The trivial region Lemma . (Lemma of []) Let V(α) = �e(αf (x,...,xm)), where f is any real function and the summation is over any finite set of values of x,...,xm. Then, for any A > , we have
LATER¶36-42
Lemma . (see [], Lemma . (Weyl’s inequality)) Suppose that ����α – a q ���� ≤ q , (a,q) = and φ(x) = αxk + αxk– + ··· + αk–x + αk. Then we have M � x= e � φ(x) � ≪ M+ε� q– + M– + qM–k�–k . Lemma . For every real number α ∈ D, we have W(α) ≪ X – δ+ ε, where W(α) = min ���G(τα) ��, ��G(τα) ��� . Proof For α ∈ D and i = ,, we choose ai, qi such that |λiα – ai/qi| ≤ q– i Q– with (ai,qi) = and ≤ qi ≤ Q. We note that aa ̸= . If q,q ≤ P, then ����aq λ λ – aq ���� ≤ ���� a/q λα qq � λα – a q ����� + ���� a/q λα qq � λα – a q ����� ≪ PQ– < q. We recall that q was chosen as the denominator of a convergent to the continued fraction for λ/λ. Thus, by Legendre’s law of best approximation, we have |q′ λ λ – a′| > q for all integers a′, q′ with ≤ q′ < q, thus |aq| ≥ q = [N–δ]. On the other hand, |aq| ≪ qqP ≪ Nδ, which is a contradiction. And so, for at least one i, P < qi ≪ Q. Hence, by Lemma ., we obtain the desired inequality for W(α). □ Lemma . The following inequality holds. � D � i= Fi(λiα)G(–α)e � – α � K (α)dα ≪ X – δ+ε. Wang and Mu Journal of Inequalities and Applications ( 2017) 2017:170 Page 8 of 9 (α)dα � ≪ � X – δ+ ε� � X + ε� � X+ ε� � X + ε� (NL) ≪ X – δ+ε from Lemmas ., . and Hölder’s inequality. □
EARLIER¶40-44
Lemma . We have � D � i= Fi(λiα)G(–α)e � – α � K (α)dα ≪ X – δ+ε. Proof By Lemmas ., . and Hölder’s inequality, we have � D � � X + ε� (NL) ≪ X – δ+ε. □ 5 The trivial region Lemma . (Lemma of []) Let V(α) = �e(αf (x,...,xm)), where f is any real function and the summation is over any finite set of values of x,...,xm. Then, for any A > , we have � |α|>A ��V(α) ��Kν(α)dα ≤ A � ∞ –∞ ��V(α) ��Kν(α)dα. Lemma . We have � c � i= Fi(λiα)G(–α)e � – α � K (α)dα ≪ X –δ+ε. Proof By Lemmas ., . and Schwarz’s inequality, we have (NL) ≪ X –δ+ε. □ 6 The proof of Theorem 1.1 From Lemmas ., . and . we conclude that J(C) ≫ X . From Lemma . it follows that J(D) = o(X ). From Lemma . we have J(c) = o(X ). Thus J ≫ X , N (X) ≫ X L–, namely, under conditions of Theorem ., ����λx + λx + λx + λx + λx + ··· + λx – p – ���� <
LATER¶42-47
(α)dα � ≪ � X – δ+ ε� � X + ε� � X+ ε� � X + ε� (NL) ≪ X – δ+ε from Lemmas ., . and Hölder’s inequality. □ 6 The trivial region Lemma . (see [], Lemma ) Let V(α) = � e � αf (x,...,xm) � , where the summation is over any finite set of values of x,...,xm and f is any real function. Then we have � |α|>A ��V(α) ��Kν(α)dα ≤ A � ∞ –∞ ��V(α) ��Kν(α)dα for any A > . The following inequality holds. Lemma . We have � c � i= Fi(λiα)G(–α)e � – α � K (α)dα ≪ X –δ+ε. Proof We have � c � i= Fi(λiα)G(–α)e � – α � K (α)dα ≪ P � +∞ –∞ ����� � i= Fi(λiα)G(–α) �����K (α)dα from Lemmas ., . and Schwarz’s inequality. □ 7 Results In this paper, we established that if λ,λ,...,λ are positive real numbers, at least one of the ratios λi/λj ( ≤ i < j ≤ ) is irrational, then the integer parts of λx + λx + λx + λx + λx + λx + λx + λx are prime infinitely often for x,x,...,x, where x,x,...,x are natural numbers. Acknowledgements The authors want to thank the reviewers for much encouragement, support, productive feedback, cautious perusal and making helpful remarks, which improved the presentation and comprehensibility of the article. This work was supported by the Higher Education Research Promotion and National Research University Project of Thailand, Office of the Higher Education Commission, through the Cluster of Research to Enhance the Quality of Basic Education.
EARLIER¶38-42
����aq λ λ – aq ���� ≤ ���� a/q λα qq � λα – a q ����� + ���� a/q λα qq � λα – a q ����� ≪ PQ– < q. We recall that q was chosen as the denominator of a convergent to the continued fraction for λ/λ. Thus, by Legendre’s law of best approximation, we have |q′ λ λ – a′| > q for all integers a′, q′ with ≤ q′ < q, thus |aq| ≥ q = [N–δ]. However, from (.) we have |aq| ≪ qqP ≪ Nδ, this is a contradiction. We have thus established that for at least one i, P < qi ≪ Q. Hence Lemma . gives the desired inequality for W(α). □ Lemma . We have � D � i= Fi(λiα)G(–α)e � – α � K (α)dα ≪ X – δ+ε. Proof By Lemmas ., . and Hölder’s inequality, we have � D � � X + ε� (NL) ≪ X – δ+ε. □ 5 The trivial region Lemma . (Lemma of []) Let V(α) = �e(αf (x,...,xm)), where f is any real function and the summation is over any finite set of values of x,...,xm. Then, for any A > , we have � |α|>A ��V(α) ��Kν(α)dα ≤ A � ∞ –∞ ��V(α) ��Kν(α)dα. Lemma . We have � c � i= Fi(λiα)G(–α)e � – α � K (α)dα ≪ X –δ+ε. Proof By Lemmas ., . and Schwarz’s inequality, we have
LATER¶42-47
(α)dα � ≪ � X – δ+ ε� � X + ε� � X+ ε� � X + ε� (NL) ≪ X – δ+ε from Lemmas ., . and Hölder’s inequality. □ 6 The trivial region Lemma . (see [], Lemma ) Let V(α) = � e � αf (x,...,xm) � , where the summation is over any finite set of values of x,...,xm and f is any real function. Then we have � |α|>A ��V(α) ��Kν(α)dα ≤ A � ∞ –∞ ��V(α) ��Kν(α)dα for any A > . The following inequality holds. Lemma . We have � c � i= Fi(λiα)G(–α)e � – α � K (α)dα ≪ X –δ+ε. Proof We have � c � i= Fi(λiα)G(–α)e � – α � K (α)dα ≪ P � +∞ –∞ ����� � i= Fi(λiα)G(–α) �����K (α)dα from Lemmas ., . and Schwarz’s inequality. □ 7 Results In this paper, we established that if λ,λ,...,λ are positive real numbers, at least one of the ratios λi/λj ( ≤ i < j ≤ ) is irrational, then the integer parts of λx + λx + λx + λx + λx + λx + λx + λx are prime infinitely often for x,x,...,x, where x,x,...,x are natural numbers. Acknowledgements The authors want to thank the reviewers for much encouragement, support, productive feedback, cautious perusal and making helpful remarks, which improved the presentation and comprehensibility of the article. This work was supported by the Higher Education Research Promotion and National Research University Project of Thailand, Office of the Higher Education Commission, through the Cluster of Research to Enhance the Quality of Basic Education.
EARLIER¶26-30
Lemma . We have � |α|>N–+δ ����� � i= fi(λiα)g(–α) �����K (α)dα ≪ X – δ. Proof It follows from Vaughan [] that for α ̸= , fi(λiα) ≪ |α|– , i = ,, fj(λjα) ≪ |α|– , j = ,, fk(λkα) ≪ |α|– , k = ,...,, g(–α) ≪ |α|–. Thus � |α|>N–+δ ����� � i= fi(λiα)g(–α) �����K (α)dα ≪ � |α|>N–+δ |α|– dα ≪ X – δ. □ Lemma . We have � +∞ –∞ � i= fi(λiα)g(–α)e � – α � K (α)dα ≫ X . Proof From (.) one has x – ···x – · max � , – ����� � i= λixi – x – ����� � dxdx ··· dx. Lai Journal of Inequalities and Applications ( 2015) 2015:357 Page 6 of 9 Let |� i= λixi – x – | ≤ , then � i= λixi – ≤ x ≤ � i= λixi – . Based on � i= λixi – > , � i= λixi – < N, one may take
LATER¶42-47
(α)dα � ≪ � X – δ+ ε� � X + ε� � X+ ε� � X + ε� (NL) ≪ X – δ+ε from Lemmas ., . and Hölder’s inequality. □ 6 The trivial region Lemma . (see [], Lemma ) Let V(α) = � e � αf (x,...,xm) � , where the summation is over any finite set of values of x,...,xm and f is any real function. Then we have � |α|>A ��V(α) ��Kν(α)dα ≤ A � ∞ –∞ ��V(α) ��Kν(α)dα for any A > . The following inequality holds. Lemma . We have � c � i= Fi(λiα)G(–α)e � – α � K (α)dα ≪ X –δ+ε. Proof We have � c � i= Fi(λiα)G(–α)e � – α � K (α)dα ≪ P � +∞ –∞ ����� � i= Fi(λiα)G(–α) �����K (α)dα from Lemmas ., . and Schwarz’s inequality. □ 7 Results In this paper, we established that if λ,λ,...,λ are positive real numbers, at least one of the ratios λi/λj ( ≤ i < j ≤ ) is irrational, then the integer parts of λx + λx + λx + λx + λx + λx + λx + λx are prime infinitely often for x,x,...,x, where x,x,...,x are natural numbers. Acknowledgements The authors want to thank the reviewers for much encouragement, support, productive feedback, cautious perusal and making helpful remarks, which improved the presentation and comprehensibility of the article. This work was supported by the Higher Education Research Promotion and National Research University Project of Thailand, Office of the Higher Education Commission, through the Cluster of Research to Enhance the Quality of Basic Education.
EARLIER¶29-33
x – ···x – · max � , – ����� � i= λixi – x – ����� � dxdx ··· dx. Lai Journal of Inequalities and Applications ( 2015) 2015:357 Page 6 of 9 Let |� i= λixi – x – | ≤ , then � i= λixi – ≤ x ≤ � i= λixi – . Based on � i= λixi – > , � i= λixi – < N, one may take λjX � � i= λi �– ≤ xj ≤ λjX � � i= λi �– , j = ,...,, hence � +∞ –∞ � i= fi(λiα)g(–α)e � – α � K (α)dα ≥ ,, � j= λj � � i= λi �– X . This completes the proof of Lemma .. □ 4 The intermediate region Lemma . We have � +∞ –∞ ��Fi(λiα) ��K (α)dα ≪ X + ε, i = ,, (.) � +∞ –∞ ��Fj(λjα) ��K (α)dα ≪ X+ ε, j = ,, (.) � +∞ –∞ ��Fk(λkα) ��K (α)dα ≪ X + ε, k = ,...,, (.)
LATER¶42-47
(α)dα � ≪ � X – δ+ ε� � X + ε� � X+ ε� � X + ε� (NL) ≪ X – δ+ε from Lemmas ., . and Hölder’s inequality. □ 6 The trivial region Lemma . (see [], Lemma ) Let V(α) = � e � αf (x,...,xm) � , where the summation is over any finite set of values of x,...,xm and f is any real function. Then we have � |α|>A ��V(α) ��Kν(α)dα ≤ A � ∞ –∞ ��V(α) ��Kν(α)dα for any A > . The following inequality holds. Lemma . We have � c � i= Fi(λiα)G(–α)e � – α � K (α)dα ≪ X –δ+ε. Proof We have � c � i= Fi(λiα)G(–α)e � – α � K (α)dα ≪ P � +∞ –∞ ����� � i= Fi(λiα)G(–α) �����K (α)dα from Lemmas ., . and Schwarz’s inequality. □ 7 Results In this paper, we established that if λ,λ,...,λ are positive real numbers, at least one of the ratios λi/λj ( ≤ i < j ≤ ) is irrational, then the integer parts of λx + λx + λx + λx + λx + λx + λx + λx are prime infinitely often for x,x,...,x, where x,x,...,x are natural numbers. Acknowledgements The authors want to thank the reviewers for much encouragement, support, productive feedback, cautious perusal and making helpful remarks, which improved the presentation and comprehensibility of the article. This work was supported by the Higher Education Research Promotion and National Research University Project of Thailand, Office of the Higher Education Commission, through the Cluster of Research to Enhance the Quality of Basic Education.
EARLIER¶3-7
Keywords: Davenport-Heilbronn method; integer variables; diophantine approximation 1 Introduction In , Brüdern et al. [] proved that if λ,...,λs are positive real numbers, λ/λ is irrational, all Dirichlet L-functions satisfy the Riemann hypothesis s ≥ k +, then the integer parts of λxk + λxk + ··· + λsxk s are prime infinitely often for natural numbers xj. Motivated by [], using the Davenport-Heilbronn method, we consider the integer part of a nonlinear form with integer variables and mixed powers , and , and establish one result as follows. Theorem . Let λ,λ,...,λ be positive real numbers, at least one of the ratios λi/λj ( ≤ i < j ≤ ) is irrational. Then the integer parts of λx + λx + λx + λx + λx + ··· + λx are prime infinitely often for natural numbers x,x,...,x. It is noted that Theorem . holds without the Riemann hypothesis. 2 Notation Throughout, we use p to denote a prime number and xj to denote a natural number. We denote by δ a sufficiently small positive number and by ε an arbitrarily small positive number. Constants, both explicit and implicit, in Landau or Vinogradov symbols may depend © 2015 Lai. This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.
LATER¶44-48
� |α|>A ��V(α) ��Kν(α)dα ≤ A � ∞ –∞ ��V(α) ��Kν(α)dα for any A > . The following inequality holds. Lemma . We have � c � i= Fi(λiα)G(–α)e � – α � K (α)dα ≪ X –δ+ε. Proof We have � c � i= Fi(λiα)G(–α)e � – α � K (α)dα ≪ P � +∞ –∞ ����� � i= Fi(λiα)G(–α) �����K (α)dα from Lemmas ., . and Schwarz’s inequality. □ 7 Results In this paper, we established that if λ,λ,...,λ are positive real numbers, at least one of the ratios λi/λj ( ≤ i < j ≤ ) is irrational, then the integer parts of λx + λx + λx + λx + λx + λx + λx + λx are prime infinitely often for x,x,...,x, where x,x,...,x are natural numbers. Acknowledgements The authors want to thank the reviewers for much encouragement, support, productive feedback, cautious perusal and making helpful remarks, which improved the presentation and comprehensibility of the article. This work was supported by the Higher Education Research Promotion and National Research University Project of Thailand, Office of the Higher Education Commission, through the Cluster of Research to Enhance the Quality of Basic Education. Competing interests The authors declare that they have no competing interests. Authors’ contributions JM drafted the manuscript. YW helped to draft the manuscript and revised the written English. Both authors read and approved the final manuscript.
EARLIER¶7-11
© 2015 Lai. This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. Lai Journal of Inequalities and Applications ( 2015) 2015:357 Page 2 of 9 on λ,λ,...,λ. We write e(x) = exp(πix). We use [x] to denote the integer part of real variable x. We take X to be the basic parameter, a large real integer. Since at least one of the ratios λi/λj ( ≤ i < j ≤ ) is irrational, without loss of generality we may assume that λ/λ is irrational. For the other cases, the only difference is in the following intermediate region, and we may deal with the same method in Section . Since λ/λ is irrational, then there are infinitely many pairs of integers q, a with |λ/λ – a/q| ≤ q–, (a,q) = , q > and a ̸= . We choose q to be large in terms of λ,λ,...,λ and
LATER¶44-48
� |α|>A ��V(α) ��Kν(α)dα ≤ A � ∞ –∞ ��V(α) ��Kν(α)dα for any A > . The following inequality holds. Lemma . We have � c � i= Fi(λiα)G(–α)e � – α � K (α)dα ≪ X –δ+ε. Proof We have � c � i= Fi(λiα)G(–α)e � – α � K (α)dα ≪ P � +∞ –∞ ����� � i= Fi(λiα)G(–α) �����K (α)dα from Lemmas ., . and Schwarz’s inequality. □ 7 Results In this paper, we established that if λ,λ,...,λ are positive real numbers, at least one of the ratios λi/λj ( ≤ i < j ≤ ) is irrational, then the integer parts of λx + λx + λx + λx + λx + λx + λx + λx are prime infinitely often for x,x,...,x, where x,x,...,x are natural numbers. Acknowledgements The authors want to thank the reviewers for much encouragement, support, productive feedback, cautious perusal and making helpful remarks, which improved the presentation and comprehensibility of the article. This work was supported by the Higher Education Research Promotion and National Research University Project of Thailand, Office of the Higher Education Commission, through the Cluster of Research to Enhance the Quality of Basic Education. Competing interests The authors declare that they have no competing interests. Authors’ contributions JM drafted the manuscript. YW helped to draft the manuscript and revised the written English. Both authors read and approved the final manuscript.
EARLIER¶3-7
Keywords: Davenport-Heilbronn method; integer variables; diophantine approximation 1 Introduction In , Brüdern et al. [] proved that if λ,...,λs are positive real numbers, λ/λ is irrational, all Dirichlet L-functions satisfy the Riemann hypothesis s ≥ k +, then the integer parts of λxk + λxk + ··· + λsxk s are prime infinitely often for natural numbers xj. Motivated by [], using the Davenport-Heilbronn method, we consider the integer part of a nonlinear form with integer variables and mixed powers , and , and establish one result as follows. Theorem . Let λ,λ,...,λ be positive real numbers, at least one of the ratios λi/λj ( ≤ i < j ≤ ) is irrational. Then the integer parts of λx + λx + λx + λx + λx + ··· + λx are prime infinitely often for natural numbers x,x,...,x. It is noted that Theorem . holds without the Riemann hypothesis. 2 Notation Throughout, we use p to denote a prime number and xj to denote a natural number. We denote by δ a sufficiently small positive number and by ε an arbitrarily small positive number. Constants, both explicit and implicit, in Landau or Vinogradov symbols may depend © 2015 Lai. This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.
LATER¶44-48
� |α|>A ��V(α) ��Kν(α)dα ≤ A � ∞ –∞ ��V(α) ��Kν(α)dα for any A > . The following inequality holds. Lemma . We have � c � i= Fi(λiα)G(–α)e � – α � K (α)dα ≪ X –δ+ε. Proof We have � c � i= Fi(λiα)G(–α)e � – α � K (α)dα ≪ P � +∞ –∞ ����� � i= Fi(λiα)G(–α) �����K (α)dα from Lemmas ., . and Schwarz’s inequality. □ 7 Results In this paper, we established that if λ,λ,...,λ are positive real numbers, at least one of the ratios λi/λj ( ≤ i < j ≤ ) is irrational, then the integer parts of λx + λx + λx + λx + λx + λx + λx + λx are prime infinitely often for x,x,...,x, where x,x,...,x are natural numbers. Acknowledgements The authors want to thank the reviewers for much encouragement, support, productive feedback, cautious perusal and making helpful remarks, which improved the presentation and comprehensibility of the article. This work was supported by the Higher Education Research Promotion and National Research University Project of Thailand, Office of the Higher Education Commission, through the Cluster of Research to Enhance the Quality of Basic Education. Competing interests The authors declare that they have no competing interests. Authors’ contributions JM drafted the manuscript. YW helped to draft the manuscript and revised the written English. Both authors read and approved the final manuscript.