On completely alternating sequences generated by rational functions
Abstract.
For any , we discuss the problem of determining completely alternating sequence of rational functions of the form , where is a polynomial of degree at-most and are distinct positive real numbers. In particular, we also prove that the sequences corresponding to the rational functions and are completely alternating if it satisfies and , respectively. Moreover, we also show a characterization of a class of joint completely monotone nets in two variables with the help of certain completely alternating sequences.
Key words and phrases:
rational function, completely monotone, Hausdorff moment2020 Mathematics Subject Classification:
Primary 44A601. Introduction
Let denote the set of nonnegative real numbers, set of nonnegative integers and set of natural numbers, respectively. For a net of real numbers and for let denotes the forward difference operator which is defined by
where stands for the -tuple with th entry equal to and elsewhere. When we refer as For let denote the operator and denote the sum Given a polynomial represents degree of A net of real numbers is said to be a joint completely monotone if
and joint completely alternating if
For we call the sequence as completely monotone sequence and completely alternating sequence if it satisfies the first and second equation, respectively. Below, we enlist a well-known characterization of completely monotone and completely alternating sequence in terms of measure (see [11, Propositions 6.11, 6.12]).
Theorem 1.1.
Let be a positive integer and be a net of real numbers.
-
is joint completely monotone if and only if it is Hausdorff moment i.e. there exists a positive Borel measure concentrated on such that
(1.1) -
The sequence is joint completely alternating if and only if there exist and a positive Borel measure concentrated on such that
(1.2) |
where represents standard inner product on
We simplify the case for the sake of this article. The sequence is completely alternating if and only if there exist and a positive Borel measure concentrated on such that
The measure in (1.1) is called the representing measure of In [13], F. Hausdorff characterized a completely monotone sequence by Hausdorff moment sequence and the above characterization about completely alternating sequence was first obtained as a special case of the Levy-Khinchin representation theory on abelian semigroups (see [9]). For a comprehensive study on these topics, we refer the reader to [9], [11], [15].
A lot of work has been done on the characterization of completely monotone sequences in terms of Hausdorff moment sequences, for instance, [10, 12, 14] and references therein. But the characterization of a completely monotone sequence induced by rational functions in terms of zeroes and poles is challenging and seems to be beyond the limit of present understanding. One of the important results in this direction is the sufficient conditions to identify completely monotone sequences introduced by Ball [7] which are solely depended on the zeroes and poles of the rational functions. This article aims to determine certain completely alternating sequence of rational functions with real zeros and poles in terms of sufficient conditions. In this article, we mainly deal with the following class of rational functions, that is, for
(1.3) |
where is a polynomial of degree at-most , and βs are distinct positive real numbers arranged in non-decreasing order.
In this article, more precisely, we concentrate on the following moment problem
Question 1.2.
Among all the sequences of rational functions , defined in (1.3), which are completely alternating?
From the operator theoretic aspect, completely monotone and completely alternating sequences are used in revealing the connection with subnormal and completely hyperexpansive operators, respectively (see [1], [2], [3], [4], [5], [6], [8]).
Before proceeding further, we encounter a non-example which shows that any sequence of rational functions is not completely alternating sequence.
Example 1.3.
Let Note that
Since is not a completely alternating sequence.
The above example motivates us to identify the completely alternating sequence among the sequence of rational functions of the form (1.3). Below, we state a result involving a set of sufficient conditions which helps to find out the completely alternating sequences induced by rational functions.
Theorem 1.4.
Let be a positive integer. Let be a polynomial given by with βs distinct positive real numbers. Let be any polynomial of degree at most . Then, the sequence is completely alternating provided
where
As an application of this result, we also develop a set of sufficient conditions to identify the completely alternating sequences induced by rational functions defined in (1.3), with and or , which solely depend on its zeros and poles (see Corollary 2.3). These sufficient conditions are very easy to check as well as turn out to be necessary in a sense discussed in RemarkΒ 2.4. In a recent article [4], the authors have studied joint complete monotonicity of the net (see [4, TheoremΒ 1.4]) and in this article we deal with this property for a more general class of two variable nets (see SectionΒ 3). We establish an equivalence between a class of completely alternating sequences and a class of joint completely monotone nets (see TheoremΒ 3.1 and RemarkΒ 3.2), which also reveals the importance of understanding completely alternating sequences.
2. completely alternating sequence
In this section, we present a proof of TheoremΒ 1.4. Along with that, we show that every member of a class of sequence of rational functions is completely alternating if it satisfies a set of conditions developed solely in terms of its zeros and poles.
We begin with the following result, which deals with rational sequences of at most degree 2 and finds out when it will be a completely alternating sequence in terms of necessary and sufficient conditions. It has a deep impact on identifying a class of completely alternating sequences.
Theorem 2.1.
Let with , then we have the following:
-
Let Then the sequence is completely alternating if and only if
-
Let Then the sequence is completely alternating if and only if
Proof.
We divide the proof into two cases. Assume that By the partial fraction decomposition,
where and Note that
In view of TheoremΒ 1.1(ii), it suffices to check that for all for the proof of both the necessity and sufficiency part. By [4, LemmaΒ 3.4], this is equivalent to and Further, it is equivalent to Now if then again by the partial fraction decomposition,
where and Note that
It is easy to verify that if and only if and Latter conditions are equivalent to and This completes the proof.
Analogous argument, as used in the first part, works here as well.
β
Next, we present the proof of one of the main Theorem of this article by generalizing the idea used in the proof of TheoremΒ 2.1.
Proof of TheoremΒ 1.4.
Note that by partial fraction decomposition, we have
(2.1) |
where and is the coefficient of in Since the range of is contained in It is easy to see using (2.1) that
where To prove the sequence is completely alternating, by TheoremΒ 1.1(ii), it is now sufficient to check that for We will use finite induction on the set For since Assume the induction hypothesis. Let If is empty, then we are done. Assume that is not empty. Let be the largest element of Let if possible Then
Since only a proper subset of the set is involved in R.H.S. so by the induction hypothesis we have
The remaining case is when We divide this case into the following two subcases.
Subcase 1.
Note that for every
The last inequality follows from the induction hypothesis.
Subcase 2.
Note that for every
The last inequality follows from the induction hypothesis. This completes the proof. β
Remark 2.2.
In TheoremΒ 1.4, turns out to be a necessary condition. Otherwise, if the sequence corresponding to the rational function with is completely alternating then the sequence turns out to be divergent, that is,
is not finite. But on the other hand, according to the TheoremΒ 1.1(ii), for a completely alternating sequence , where is a real number as in (1.2).
As an application to TheoremΒ 1.4, we obtain the following classes of completely alternating sequences.
Corollary 2.3.
Let and be any of the following
-
with
-
with
Then the sequence is completely alternating.
Proof.
(i) Note that
where This can be rewritten as
(we assume that the product over the empty set is ). Under the given conditions, it is easy to see that By TheoremΒ 1.4, the sequence is completely alternating. A similar proof works for (ii). β
Remark 2.4.
The sufficient conditions in CorollaryΒ 2.3 turn out to be necessary if we further assume that, in the first case where degree of is same as degree of , the sequences for each and are complete alternating and in the second case where degree of is bigger than degree of by 1, the sequences for each are completely alternating. The proof follows from the TheoremΒ 2.1.
3. A link between CA sequence and JCM net
In this section, we establish an equivalence between completely alternating sequences and two variable joint completely monotone nets and show an immediate application of the CorollaryΒ 2.3.
In [5, PropositionΒ 6] (also see [12, TheoremΒ 2.5]), the authors have shown a correspondence between completely alternating sequences and completely monotone sequences. In the following result, we establish a relation between completely alternating sequences and two variable joint completely monotone nets. This also extends [4, TheoremΒ 1.4(i)].
Theorem 3.1.
Let and be sequences of positive real numbers. Then the following are equivalent
-
the sequence is completely alternating,
-
is a joint completely monotone net.
Moreover, if holds and is a completely monotone sequence, then the sequence is a joint completely monotone net.
Proof.
To prove the equivalence, we note the following identity:
(3.1) |
(i)(ii): Assume that is a completely alternating sequence. By [11, PropositionΒ 6.10], is a completely monotone sequence, for every Equivalently, the sequence is completely monotone. This together with the fact that is completely monotone, yields that for every
Combining this with (3.1), we get
This shows that is a joint completely monotone net.
(ii)(i): Assume that is a joint completely monotone net. This together with (3.1) yields that for every
(3.2) |
Since the integrand in (3.2) is a continuous function in
This yields that for every is a completely monotone sequence and equivalently, is a completely alternating sequence. To see the moreover part, assume that holds and is a completely monotone sequence. Since the product of completely monotone net is completely monotone, the sequence is a joint completely monotone net. This completes the proof. β
Remark 3.2.
We intuitively believe that it might be possible to relax the condition, that is, the sequence is also completely monotone, in TheoremΒ 3.1 but we do not know any proof of it. Therefore, we do not have any answer to the following question, which is a generalization of TheoremΒ 1.4 in [4].
Question 3.3.
Let and be sequences of positive real numbers. Is it true that the following statements are equivalent
-
the sequence is completely alternating,
-
is a joint completely monotone net?
Acknowledgements: The authors are grateful to Prof. Sameer Chavan for some insightful comments/suggestions. The second named author acknowledges the Department of Mathematics, Birla Institute of Technology and Science K.K. Birla Goa Campus for warm hospitality and is grateful to the DST-INSPIRE Faculty Fellowship. The first-named authorβs research is supported by the DST-INSPIRE Faculty Fellowship No. DST/INSPIRE/04/2020/001250.
References
- [1] A. Anand, S. Chavan, A moment problem and joint q-isometry tuples. Complex Anal. Oper. Theory 11 (2017), 785β810.
- [2] A. Anand, S. Chavan, Module tensor product of subnormal modules need not be subnormal. J. Funct. Anal. 272(11) (2017), 4752β4761.
- [3] A. Anand, S. Chavan, Z. J. JabΕoΕski, J. Stochel, A solution to the Cauchy dual subnormality problem for 2-isometries, J. Funct. Anal. 277 (2019), 108292, 51 pp.
- [4] A. Anand, S. Chavan, R. Nailwal, Joint complete monotonicity of rational functions in two variables and toral m-isometric pairs, J. Operator Theory, to appear, (2023).
- [5] A. Athavale, On completely hyperexpansive operators, Proc. Amer. Math. Soc. 124 (1996), 3745β3752.
- [6] A. Athavale, A. Ranjekar, Bernstein functions, Complete hyperexpansivity and subnormality I, Integral Equations Operator Theory 43 (2002), 253-263.
- [7] K. Ball, Completely monotone rational functions and Hallβs marriage theorem, J. Comb. Theory 61 (1994), 118-124.
- [8] C. Benhida and R. E. Curto and G. R. Exner, Moment infinite divisibility of weighted shifts: sequence conditions, Complex Anal. Oper. Theory 16 (2022), 23 pp.
- [9] C. Berg, J. P. R. Christensen, P. Ressel, Positive definite functions on abelian semigroups, Math. Ann. 223 (1976), 253-274.
- [10] C. Berg, A. DurΓ‘n, Some transformations of Hausdorff moment sequences and harmonic numbers, Canad. J. Math. 57 (2005), 941β960.
- [11] C. Berg, J. P. R. Christensen, P. Ressel, Harmonic analysis on semigroups. Theory of positive definite and related functions. Graduate Texts in Mathematics, 100. Springer-Verlag, New York, 1984. x+289 pp.
- [12] S. Ghara, S. Kumar, On the sum of two subnormal kernels, Journal of Mathematical Analysis and Applications 469(2) (2019), 1015-1027
- [13] F. Hausdorff, Momentprobleme fΓΌr ein endliches intervall, Math. Z. 16(1) (1923), 220β248
- [14] Md. R. Reza, G. Zhang, Hausdorff moment sequences induced by rational functions, Complex Anal. Oper. Theory 13 (2019), 4117-4142.
- [15] D. V. Widder, The Laplace Transform, Princeton Mathematical Series, vol. 6. Princeton University Press, Princeton, N. J., 1941. x+406 pp.