Open Access

Some Inequalities for Modified Bessel Functions

Journal of Inequalities and Applications20102010:253035

DOI: 10.1155/2010/253035

Received: 15 October 2009

Accepted: 28 December 2009

Published: 24 January 2010

Abstract

We denote by and the Bessel functions of the first and third kinds, respectively. Motivated by the relevance of the function , , in many contexts of applied mathematics and, in particular, in some elasticity problems Simpson and Spector (1984), we establish new inequalities for . The results are based on the recurrence relations for and and the Turán-type inequalities for such functions. Similar investigations are developed to establish new inequalities for .

1. Introduction

Inequalities for modified Bessel functions and have been established by many authors. For example, Bordelon [1] and Ross [2] proved the bounds

(1.1)

The lower bound was also proved by Laforgia [3] for larger domain . In [3] also the following bounds:

(1.2)
(1.3)
(1.4)

have been established; see also [4]

In this paper we continue our investigations on new inequalities for and , but now our results refer not only to a function or at two different points and , as in (1.1)–(1.4), but to two functions and ( and ) and, more precisely, to the ratio . This kind of ratios appears often in applied sciences. Recently, for example, the ratio has been used by Baricz to prove an important lemma (see [5, Lemma 1]) which provides new lower and upper bounds for the generalized Marcum -function
(1.5)

(see also [6]). This generalized function and the classical one, , are widely used in the electronic field, in particular in radar communications [7, 8] and in error performance analysis of multichannel dealing with partially coherent, differentially coherent, and noncoherent detections over fading channels [7, 9, 10].

The results obtained in this paper are proved as consequence of the recurrence relations [11, page 376; 9.6.26]

(1.6)
(1.7)

and the Turán-type inequalities

(1.8)
(1.9)

proved in [12, 13], respectively (see also [14] for (1.9)). Inequalities (1.8)-(1.9) have been used, recently, by Baricz in [15], to prove, in different way, the known inequalities

(1.10)
(1.11)

The results are given by the following theorems.

Theorem 1.1.

For real let be the modified Bessel function of the first kind and order . Then
(1.12)

In particular, for , the inequality holds also true.

Theorem.

For real let be the modified Bessel function of the third kind and order . Then
(1.13)

In particular, for , the inequality holds also true.

2. The Proofs

Proof of Theorem 1.1.

The upper bound for the ratio follows from the inequality
(2.1)

proved by Soni for [16], and extended by Näsell to [17].

To prove the lower bound in (1.12), we substitute the function given by (1.6) in the Turán-type inequality (1.8). We get, for ,

(2.2)
that is,
(2.3)
We denote by and observe that for , by (2.1), . With this notation (2.3) can be written as
(2.4)
which gives, for ,
(2.5)
that is,
(2.6)

which is the desired result.

Remark.

For , Jones [18] proved stronger result than (2.1) that the function decreases with respect to , when .

Proof of Theorem 1.2.

The proof is similar to the one used to prove Theorem 1.1. By
(2.7)

we get , for .

We substitute the function given by (1.7) in (1.9). We get

(2.8)
or, equivalently
(2.9)
that is,
(2.10)
Finally, we obtain
(2.11)

which is the desired result (1.13).

Remark.

By means the integral formula [11, page 181]

(2.12)
follows immediately the inequality
(2.13)
and consequently
(2.14)

Since when , only in this case the above upper bound for improves the (1.13) one.

Remark.

We observe that by Theorem 1.1 we obtain an upper bound for the function , . The investigations of the properties of are motivated by some problems of finite elasticity [19, 20]. By (1.12) we find
(2.15)

in particular, for , we also have .

3. Numerical Considerations

Baricz obtained, for each , the following similar lower bound for the ratio (see [5, formula (5)])

(3.1)

where is the unique simple positive root of the equation . Inequality (3.1) is reversed when . It is possible to prove that, for , our lower bound in (1.12) for the ratio provides an improvement of (3.1).

Proposition.

Let be . Putting and , one has , for all .

Proof.

From the inequality we obtain, by simple calculations, the following one which is satisfied for all when .

We report here some numerical experiments, computed by using mathematica.

Example.

In the first case we assume . In Figure 1 we report the graphics of the functions (solid line) and the respective lower bounds (short dashed line) and (long dashed line) on the interval .

In Table 1 we report also the respective numerical values of the differences and in some points .

Table 1

 

Figure 1

Remark.

By some numerical experiments we can conjecture that the lower bound (3.1) holds true also when and, in particular, for these values of we have . See, for example, in Figure 2 the graphics of the functions (solid line) and the respective lower bounds (short dashed line) and (long dashed line) on the interval when .

Figure 2

Example.

In this case we assume , then we report, in Figure 3, the graphics of the functions (solid line) and the respective lower bounds (short dashed line) on the interval .In Table 3 we report also the respective numerical values of the differences in some points .

Figure 3

Example.

Also in this case we assume . In Figure 4 we report the graphics of the functions (solid line) and the respective upper bound (short dashed line) on the interval .In Table 2, we report also the respective numerical values of the difference in some points .

Table 2

 

Table 3

 

Figure 4

Example.

In this last case we assume . In Figure 5 we report the graphics of the functions (solid line) and the respective upper bound (short dashed line) on the interval . In Table 4 we report also the respective numerical values of the difference in some points .

Table 4

 

Figure 5

Remark.

We conclude this paper observing that, dividing by inequalities (1.10)-(1.11) and integrating them from to ( ), we obtain the following new lower bounds for the ratios and :
(3.2)
(3.3)

For a survey on inequalities of the type (3.2) and (3.3) see [4].

In the following Tables 5 and 6 we confront the lower bounds (1.1)–(3.2) and (1.4)–(3.3), respectively, for different values of in the particular cases and . Let

Table 5

 

Table 6

 

(3.4)

then we have Tables 5 and 6.

By the values reported on Table 5 it seems that is a lower bound much more stringent with respect to for every (moreover we recall that (3.2) holds true also for ), while by the values reported on Table 6 it seems that is a lower bound more stringent with respect to for (but we recall that (3.3) holds true also for and ).

Declarations

Acknowledgment

This work was sponsored by Ministero dell'Universitá e della Ricerca Scientifica Grant no. 2006090295.

Authors’ Affiliations

(1)
Department of Mathematics, Rome Tre University

References

  1. Bordelon DJ: Solution to problem 72–15. SIAM Review 1973, 15: 666–668.MathSciNetGoogle Scholar
  2. Ross DK: Solution to problem 72–15. SIAM Review 1973, 15: 668–670.Google Scholar
  3. Laforgia A: Bounds for modified Bessel functions. Journal of Computational and Applied Mathematics 1991, 34(3):263–267. 10.1016/0377-0427(91)90087-ZMATHMathSciNetView ArticleGoogle Scholar
  4. Baricz Á: Bounds for modified Bessel functions of the first and second kind. Proceedings of the Edinburgh Mathematical Society. In press Proceedings of the Edinburgh Mathematical Society. In press
  5. Baricz Á: Tight bounds for the generalized Marcum -function. Journal of Mathematical Analysis and Applications 2009, 360(1):265–277. 10.1016/j.jmaa.2009.06.055MATHMathSciNetView ArticleGoogle Scholar
  6. Baricz Á, Sun Y: New bounds for the generalized marcum -function. IEEE Transactions on Information Theory 2009, 55(7):3091–3100.MathSciNetView ArticleGoogle Scholar
  7. Marcum JI: A statistical theory of target detection by pulsed radar. IRE Transactions on Information Theory 1960, 6: 59–267. 10.1109/TIT.1960.1057560MathSciNetView ArticleGoogle Scholar
  8. Marcum JI, Swerling P: Studies of target detection by pulsed radar. IEEE Transactions on Information Theory 1960, 6: 227–228.View ArticleGoogle Scholar
  9. Nuttall AH: Some integrals involving the function. IEEE Transactions on Information Theory 1975, 21: 95–96. 10.1109/TIT.1975.1055327MATHMathSciNetView ArticleGoogle Scholar
  10. Simon MK, Alouini MS: Digital Communication over Fadding Channels: A Unified Approach to Performance Analysis. John Wiley & Sons, New York, NY, USA; 2000.View ArticleGoogle Scholar
  11. Abramowitz M, Stegun IA: Handbook of Mathematical Functions with Formulas, Graphs, and Mathematical Tables, National Bureau of Standards Applied Mathematics Series. Volume 55. U.S. Government Printing Office, Washington, DC, USA; 1964:xiv+1046.Google Scholar
  12. Lorch L: Monotonicity of the zeros of a cross product of Bessel functions. Methods and Applications of Analysis 1994, 1(1):75–80.MATHMathSciNetGoogle Scholar
  13. Ismail MEH, Muldoon ME: Monotonicity of the zeros of a cross-product of Bessel functions. SIAM Journal on Mathematical Analysis 1978, 9(4):759–767. 10.1137/0509055MATHMathSciNetView ArticleGoogle Scholar
  14. Laforgia A, Natalini P: On some Turán-type inequalities. Journal of Inequalities and Applications 2006, 2006:-6.Google Scholar
  15. Baricz Á: On a product of modified Bessel functions. Proceedings of the American Mathematical Society 2009, 137(1):189–193.MATHMathSciNetView ArticleGoogle Scholar
  16. Soni RP: On an inequality for modified Bessel functions. Journal of Mathematical Physics 1965, 44: 406–407.MATHMathSciNetView ArticleGoogle Scholar
  17. Näsell I: Inequalities for modified Bessel functions. Mathematics of Computation 1974, 28: 253–256.MathSciNetView ArticleGoogle Scholar
  18. Jones AL: An extension of an inequality involving modified Bessel functions. Journal of Mathematical Physics 1968, 47: 220–221.MATHView ArticleGoogle Scholar
  19. Simpson HC, Spector SJ: Some monotonicity results for ratios of modified Bessel functions. Quarterly of Applied Mathematics 1984, 42(1):95–98.MATHMathSciNetGoogle Scholar
  20. Simpson HC, Spector SJ: On barrelling for a special material in finite elasticity. Quarterly of Applied Mathematics 1984, 42(1):99–111.MATHMathSciNetGoogle Scholar

Copyright

© A. Laforgia and P. Natalini. 2010

This article is published under license to BioMed Central Ltd. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.