Upload
others
View
1
Download
0
Embed Size (px)
Citation preview
Internat. J. Math. & Math. Scl.VOL. 16 NO. 2 (1993) 255-258
255
REMARKS ON ORTHOGONAL POLYNOMIALS WITH RESPECT TOVARYING MEASURES AND RELATED PROBLEMS
XlN LI
Department of MathematicsUniversity of Central Florida
Orlando, FL 32816
(Received March 26, 1992 and in revised form May 4, 1992)
ABSTRACT. We point out the relation between the orthogonal polynomials with respect to
(w.r.t.) varying measures and the so-called orthogonal rationals on the unit circle in the complex
plane. This observation enables us to combine different techniques in the study of these
polynomials and rationals. As an example, we present a simple and short proof for a known result
on the weak-star convergence of orthogonal polynomials w.r.t, varying measures. Some related
problems are also considered.
KEY WORDS AND PHRASES. Orthogonal polynomial and rational, weak-star convergence.
1991 AMS SUBJECT CLASSIFICATION CODES. Primary 42A05, Secondary 33C45, 41A28.
1. INTRODUCTION.Let a be a finite positive measure on the unit circle r: {zECI zl 1}, and let
w,(z) zn+.. be a sequence of polynomials whose zeros {z,,,}’= all lie in the closed unit disk
zl _< 1. Assume
r ,,,(z) 12 < ’then, for each n, we can consider the orthonormal polynomials ,,,m(z)=a,.,z"+---(a,,,>O)w.r.t, d/I w,,(z) 12, i.e., polynomials satisfying
and
] z-J,n,m{z) =0, j=0,1 m,r w,(z)
f i,,.,(z)ld,
2 r iw,(z)l_X.The sequences {,,,(z)}= 0 are called orthogonal polynomials w.r.t, varying measures der/Iw,(z)
(n 1,2 ). They appear in the study of simultaneous Pad approximation and related problems.
The following weak-star convergence result plays an important role in the study of convergence of
{,,,()}= -
256 X. LI
THEOREM 1. (Lopez, [8])limn._.,oo n= 1( z,, l) o, then
in the weak-star topology.
If the zeros of w.(z) satisfy the condition
One nice thing about this result is that it holds without any condition on the finite positive
measure da. We will present a simple and short proof of Theorem 1. But before we get to the
proof, we remark that the so-called orthogonal rationals (with prescribed poles) are also of current
interest, especially in problems related to electrical engineering. (See, e.g., [2, 3, 4, 5]) These
applications were more concerned with the algebraic properties of orthogonal rationals. It turns out
that orthogonal rationals are just special case of orthogonal polynomials w.r.t, varying measures. It
is the main purpose of this note to point out this relation. To see this relation, let us recall the
definition of orthogonal rationals with respect to
Let {zi},= be points in zl _< 1, and denote O,(z): l-I’= 10- z,).DEFINITION 2. (cf. [2]) For n 1,2, the rationals P,(z) with P,(z) a,z" + (a,, > O)
satisfying
(z).for r(z) tQn{z), q 9n- 1}, and -/ F Q-oPn(z) ’-(z)do. =0
rdo.=l,
are called orthogonal rationals w.r.t.
Now, from Definition 2, it can be seen that with Q,(z) treated as w,(z), the corresponding
.,,(z) is P,(z). So orthogonal rationals are indeed special case of orthogonal polynomials w.r.t.
varying measures.
We go back to consider orthogonal polynomials w.r.t, varying measures. The following result,
which is essentially proved in the orthogonal rational setting (cf. e.g., [2]), will help us to simplify
the proof of Theorem 1 in [8].LEMMA 3. If r and s belong to %(w,): {p(z)/w.(z): p ,,}, then
/+/-2 rr(z)7 .,.()1 dz rr(Z)S(z)do."
PROOF. By [6, formula (1.20)], we have
r I,,,,(z)l ldzl r w,(z)
for all p, q 9n. So,
/ p(z) (q(z) Wn(Z) 12 2j./ p(z) (q(z) do’,2-7 r K,o.(z)/I I lagr \w.(z)J
which implies the lemma.
REMARKS. (i) Taking r in Lemma 3 gives us
do..r ,,,(z) dzl r
(1)
ORTHOGONAL POLYNOMIALS WITH RESPECT TO VARYING MEASURES 257
(ii) Lemma 3 is equivalent to the following:
/ z-"t’2"(z) w"(z) 2J/ z-"t’2"(z)do’,Idzl r Iw,(z)[
(2)
for all P2. E 92,,.We also need the following result about weighted approximation with varying weights in
Loo-norm.LEMMA 4. The space
T.: iw.(z)12"P2. E2,,
is dense in c(r) in the L-norm if and only if
lira (1-
PROOF. This is a simple generalization of some known results (see, e.g., [1, p. 244] and [2,Theorem 8..2]). We only mention the following fact (cf. [1, p. 243])
inf II z,.It) ( P2n + l’I ,n,, # o(Z Zn, ,)(Z 1/Z--,;) z.,, #o
PROOF OF THEOREM 1. Now the proof of Theorem 1 follows from (1), (2), and Lemma 4
by using a standard argument (cf. e.g., [9, p. 248]).
Recently, in the study of frequency analysis (see, e.g., [7]), the following orthogonalpolynomials w.r.t, varying measures appear naturally, i.e., polynomials
+ (3,.,, > 0) satisfying
l___f z-,g,n.(z) lwn(z)O2da=O,j=O, mlF
and
/ . ,.(z)w.(z)l 2da 1.2"j F
For convenience, let us assume that the zeros of w,(z) all lie outside the unit circle r, i.e., z,.il > 1.
The asymptotics of ,,.,,(z)w,,(z) are wanted in application ([7]). Here we will not discuss this
problem but instead point out the relation of these polynomials {,.,,} with the following ones
{,,.,,}: Assume An. > 0(1 < <_ n,n >_ 1) are given, then define polynomials ,.,,(z)+ (7,.,, > 0) satisfying
r i=1
for j 0,1 m + n 1, and
(z) 12da + A,, ,. ,,(z,. i) 12 .F i=1
The polynomials {#,.,,} 0 are the orthonormal polynomials w.r.t, measure da/(2,)
+ ’= 1A,.z, having finitely many mass points off the unit circle r (where $, denotes the point.,unit measure with support at z). We have
258 X. LI
THEOREM 5. For each n fixed, there hold
lim ,,,,,(z) ,,,,,,(z)w,,(z),
l<i<n
locally uniformly in C, where m 0,1,2,
PROOF. We use a normal family argument. For n and m chosen, since f rl,,,(z) lZda _< 2,
then the set : {n,m:An,,>O, <_i <_n} as a subset of finite dimensional normed space is a
uniformly bounded set. Thus as a set of analytic functions (polynomials) is a normal family over
C. Let f(z)E 9,+,, be a limit of a subsequence O’c O, then since [’__IA,,, J,,,,,(z,,,)J <_ 1, we
have
lim
’,, ,(z,, ,) 0f(z,,,)
6,,,,, e
for i= 1,2 n. So 1’(z)=wn(z)g(z with some g(z)eg,. Now we just need to use the extremal
property of (monic) orthogonal polynomials to assert that g(z)= Cn,,Jz). This completes the proofof Theorem 5.
REFERENCES
1. ACHIESER, N.I., Theory of Approximation, Frederick Ungar Publ. Co., New York, 1956.
2. BULTHEEL, A., GONZALEZ-VERA, P., HENDRIKSEN, E., & NJASTAD, O., A Szegtltheory for rational functions, manuscript, 1991.
3. DEWILDE, P., & DYM, H., Lossless inverse scattering, digital filters, and estimation theory,IEEE Trans. Inform. Theory, Vol. IT-30 (1984), 644-663.
4. DEWILDE, P., & DYM, H., Schur recursions, error formulas, and convergence of rationalestimators for stationary stochastic sequences, IEEE Tran. Inform. Theory, Vol. IT-27(198I), 446-461
5. DJRBASHIAN, M.M., A survey on the theory of orthogonal systems and some openproblems, Orthogonal Polynomials: Theory and Applications, (P. Nevai ed.), KluwerAcad. Publ., Boston, 1990, 135-146.
6. GERONIMUS, YA., Orthogonal Polynomials. Consultants Bureau, New York, 1961.
7. JONES, W.B., NJASTAD, O., & SAFF, E.B., Szeg6 polynomials associated with Wiener-Levinson filters, J. Comp. Appl. Math. 32 (1990), 387-406.
8. LOPEZ LAGOMASINO, G., Asymptotics of polynomials orthogonal with respect to varyingmeasures, Constr. Approx. 5 (1989), 199-219.
9. MT, A., NEVAI, P. & TOTIK, V., Strong and weak convergence of orthogonalpolynomials, Amer. J. Math. 109 (1987), 239-282.
Submit your manuscripts athttp://www.hindawi.com
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
MathematicsJournal of
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Mathematical Problems in Engineering
Hindawi Publishing Corporationhttp://www.hindawi.com
Differential EquationsInternational Journal of
Volume 2014
Applied MathematicsJournal of
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Probability and StatisticsHindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Journal of
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Mathematical PhysicsAdvances in
Complex AnalysisJournal of
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
OptimizationJournal of
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
CombinatoricsHindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
International Journal of
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Operations ResearchAdvances in
Journal of
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Function Spaces
Abstract and Applied AnalysisHindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
International Journal of Mathematics and Mathematical Sciences
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
The Scientific World JournalHindawi Publishing Corporation http://www.hindawi.com Volume 2014
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Algebra
Discrete Dynamics in Nature and Society
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Decision SciencesAdvances in
Discrete MathematicsJournal of
Hindawi Publishing Corporationhttp://www.hindawi.com
Volume 2014 Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Stochastic AnalysisInternational Journal of