Научная статья на тему 'Modification of the Mishou Theorem'

Modification of the Mishou Theorem Текст научной статьи по специальности «Математика»

CC BY
65
14
i Надоели баннеры? Вы всегда можете отключить рекламу.
Журнал
Чебышевский сборник
Scopus
ВАК
RSCI
Область наук
Ключевые слова
ДЗЕТА-ФУНКЦИЯ ГУРВИЦА / ДЗЕТА-ФУНКЦИЯ РИМАНА / ПРОСТРАНСТВО АНАЛИТИЧЕСКИХ ФУНКЦИЙ / УНИВЕРСАЛЬНОСТЬ / HURWITZ ZETA-FUNCTION / RIEMANN ZETA-FUNCTION / SPACE OF ANALYTIC FUNCTIONS / UNIVERSALITY
i Надоели баннеры? Вы всегда можете отключить рекламу.
iНе можете найти то, что вам нужно? Попробуйте сервис подбора литературы.
i Надоели баннеры? Вы всегда можете отключить рекламу.

Текст научной работы на тему «Modification of the Mishou Theorem»

ЧЕБЫШЕВСКИИ СБОРНИК

Том 17 Выпуск 3

УДК 519.14

МОДИФИКАЦИЯ ТЕОРЕМЫ МИШУ

А. Лауринчикае, (г, Вильнюс, Литва), Л, Мешка (г, Вильнюс, Литва)

Аннотация

В 2007 г. Г. Мишу доказал совместную теорему унивурсальности для дзета-функции Римана С (в) и дзета-функции Гу рвица С (в, а) с трансцендентным параметром а об одновременном приближении пары функций из широкого класса аналитических функций сдвигами (С (в + гт), £ (в + гт, а)), т € М. Он получил, что множество таких сдвигов, приближающих данную пару аналитических функций, имеет положительную нижнюю плотность. В статье получено, что множество таких сдвигов имеет положительную плотность для всех е > 0 за исключением счетного множеств а значений е, где е — точность приближения.

Результаты аналогичного типа также получены для сложных функций Г ( £ (в), С (в, а)) для некоторых классов операторов Г в пространстве аналитических функций.

Ключевые слова: дзета-функция Гурвица, дзета-функция Римана, пространство аналитических функций, универсальность.

Библиография: 21 названий.

The Mishou theorem asserts that a pair of analytic functions from a wide class can be approximated by shifts of the Riemann zeta and Hurwitz zeta-functions (Z(s + ¿t), Z(s + ¿t, a)) with transcendental a, t g R, and that the set of such t has a positive lower density. In the paper, we prove that the above set has a positive density for all but at most countably many e > 0, where £ is the accuracy of approximation. We also obtain similar results for composite functions F(Z(s),Z(s, a)) for some classes of operator F.

Keywords: Hurwitz zeta-function, Riemann zeta-function, space of analytic functions, universality.

Bibliography: 21 titles.

1. Introduction

Let Z(s), s = a + it, be the Riemann zeta-function. In 1975, S. M. Voronin discovered [21] the universality property of Z(s) which means that a wide class of non-vanishing analytic functions can be approximated by shifts Z(s + ¿t), t G R. The non-vanishing of approximated functions is

Z(s)

Now let 0 < a ^ 1 be a fixed parameter, and Z(s, a) denotes the Hurwitz zeta-function which is defined, for a > 1, by the series

MODIFICATION OF THE MISHOU THEOREM

A. Laurincikas, (Vilnius, Lithuania), L. Meska (Vilnius, Lithuania)

Abstract

and can be meromorphically continued to the whole complex plane. Clearly, ((s, 1) = ((s), and

((s, 1) =(2s - 1)Z(s).

For other values of the parameter a, the function ((s, a) has no Euler product. It is well known that the Hurwitz zeta-function with transcendental or rational = 1, \ parameter a is also universal in the above sense, however, its shifts ((s + ir, a) approximate not necessarily non-vanishing analytic functions. The universality of ((s, a) with algebraic irrational a is an open problem.

Some other zeta-functions are also universal in the Voronin sense. The universality for zeta-functions of certain cusp forms was obtained in [12], for periodic zeta-functions was studied in [20] and [15], while the works [2], [4] and [5] are devoted to periodic Hurwitz zeta-functions. Universality theorems for Lerch zeta-functions can be found in [11]. A very good survey on universality of zeta-functions is given in [16].

In [19], H. Mishou began to study the so-called mixed joint universality. In this case, a collection of analytic functions are simultaneously approximated by shifts of a collection of zeta-functions consisting from functions having the Euler product and having no such a product. H. Mishou considered the pair (((s), ((s, a)) with transcendental a. For the statement of the Mishou theorem, we need some notation. Let D = {s e C : i <ct< l}. Denote by K the class of compact subsets of the strip D with connected complements. Moreover, let H(K), K e K, be the class of continuous functions on K which are analytic in the interior of K, and let H0(K), K e K, be the subclass of H(K) consisting from non-vanishing functions on K. Denote by measA the Lebesgue measure of a measurable set A C R. Then H. Mishou proved [19] the following theorem.

Theorem 1. Suppose that a is transcendental number. Let Ki,K2 e K, and /i(s) e H0(Ki); /2(s) e H(K2). Then, for every e > 0

liminf ^measjr e [0; T] : sup |((s + ir) — /i(s)| < e,

T I seKl

sup |Z(s + ir, a) — /2(s)| < e ^ > 0.

sK

seK 2

Mixed joint universality theorems are also proved in [3], [7] and [10].

lim inf lim Z(s)

done in [13] and [18], and, in the case of ((s, a) a similar theorem was obtained in [14]. Let P be the set of all prime numbers, No = N U {0}, and

L(a, P) = {(log(m + a) : m e N0), (logp : p e P)} .

Theorem 2. Suppose that the set L(a, P) is linearly independent over the field of rational numbers Q. Let Ki,K2 e K , and /i(s) e H0(Ki) /2(s) e H(K2). Then the limit

lim ^measj r e [0; T] : sup |( (s + ir) — /i(s)| < e, sup |( (s + ir, a) — /2(s)| < ^ > 0

T ^ s€K1 s€K2 J

e > 0

For example, if a is transcendental, then the set L(a, P) is linearly independent over Q.

Let H(G) be the space of analytic functions on G equipped with the topology of uniform convergence on compacta. In [9], universality theorems were proved for the functions F(((s), ((s, a)) with some operators F : H2(D) ^ H(D). Let

S = {g e H(D) : g(s) = 0 or g(s) = 0} . Then, for example in [9], the following assertion was obtained.

Theorem 3. Suppose that a is transcendental, and that F : H2(D) ^ H(D) is a continuous operator such that, for every open set G C H(D) the set (F-1G) n (S x H(D)) is non-empty. Let K G K and f (s) G H(D). Then, for every e > 0,

liminf —meas< t G [0; T] : sup |F(Z(s + iT), Z(s + iT, a)) - f (s)| < 4 > 0. t^^ T I seK J

More general results are obtained in [10].

Clearly, the transcendence of a in Theorem 3 can be replaced by a linear independence over Q of the set L(a, P). Therefore, we will prove the following theorem.

Theorem 4. Suppose that the set L(a, P) is linearly independent over Q, and that F, K and f(s)

lim ¿measi t G [0; T] : sup |F (Z (s + iT ),Z (s + iT,a)) - f (s)| < e) > 0 (1)

T I s£K J

e > 0

Now, let V be an arbitrary positive number, Dy = {s e C : 2 < ^ < 1, |t| < V} and

Sv = {g G H(Dv) : g(s) = 0 or g(s) = 0} . For brevity, we use the notation H2(DV, D) = H(DV) x H(D).

Theorem 5. Suppose that the set L(a, P) is linearly independe nt over Q, and th at K and f (s) are the same as in Theorem 3, and V > 0 is such that K C Dy- Let F : H2(Dy, D) ^ H(Dy) 6e a continuous operator such that , for each polynomial p = p(s), iAe sei (F-1{p}) n (Sy x H(Dy))

e > 0

For example, Theorem 5 implies the modified universality of the functions

c1Z(s) + c2Z(s, a^d c1Z'(s) + c2Z'(s, a) with c1, c2 e C \ {0}. Let a1,..., ar be arbitrary distinct complex numbers, and

Hai,...,ar(D) = {g G H(D) : (g(s) - aj)-1 G H(D), j = 1,..., r} .

Theorem 6. Suppose that the set L(a, P) is linearly independent over Q, and F : H2(D) ^ H(D) is a continuous operator such that F(S x H(D)) D Hai,...,ar(D^. Wften r = 1, lei K G K,

f(s) G H(K) f(s) = a1 K e > 0. // r ^ 2 K C D ¿s an arbitrary compact subset, and f (s) G Hai,...,ar(D); iften ifte Hmfi (1)

e > 0

The case r = 1 with a1 = 0 shows that, for F(g1(s),g2(s)) = egl(s)+g2the limit (1) exists for all but at most countablv many e > 0. If r = 2 and a1 = 1 a2 = —1, then, for example, for F(g1(s),g2(s)) = cos(g1(s) + g2(s)^d f(s) G H1)-1(D), the limit (1) exists for all but at most e > 0

Theorem 7. Suppose that the set L(a, P) is linearly independe nt over Q, F : H 2(D) ^ H (D) ¿s a continuous operator, K C D is a compact subset, and f (s) G F(S x H(D)). Tften the limit (1)

e > 0

2. Lemmas

In this section, we present probabilistic theorems on the weak convergence of probability measures in the space of analytic functions. Let y = {s e C : |s| = 1}, and

<x

Qi = n Yp and Q2 = n Ym,

p m=0

where yp = Y f°r aU P e ^^d Ym = Y f°r aU m e N0. By the Tikhonov theorem, the tori Qi and Q2 with the product topology and operation of pointwise multiplication are compact topological Abelian groups. Similarly, Q = Qi x Q2 is also a compact topological Abelian group. Therefore, denoting by B(X) the Borel a-field of the space X, we have that, on (Q,B(Q)), the probability Haar measure mH can be defined, and we obtain the probability space (Q, B(Q),mH). Denote by wi(p^d w2(m) the projections of wi e Qi and w2 e Q2 to the coordinate spaces yp P e P, and Ym, m e N0, respectively, and, on the probability space (Q, B(Q),mH), define the H2(D)-valued random element ((s, w), w = (wi,w2) e Q, by the formula

((s, a, w) = (((s, wi), ((s, a, W2)),

where

z c.wi) = n (1 - "

p

and

N ^ W2(m) Z (s,a,W2) = > 7-■—-.

^ (m + a)s

m=0

Moreover, let

Pz(A) = mH (w e Q : ((s,a,w) e A) , A e B(H2(D)), i.e., Pz is the distribution of the random element ((s,w). We set ((s, a) = (((s), ((s, a)), and

Pt(A) d=f ^meas {r e [0,T] : ((s + ir, a) e A} , A e B(H2(D)).

Lemma 1. Suppose that the set L(a, P) is linearly independent over Q. Then P converges weakly to Pz as T ^ to.

Proof. The lemma for transcendental a is proved in [19], Theorem 1, however, the transcendence of a is used only for the linear independence of the set L(a, P).

Let Xi and X2 be two metric spaces, and let the function u : Xi ^ X2 be (B(Xi), B(X2))-measurable. Then every probability measure P on (Xi, B(Xi)) induces on (X2, B(X2)) the unique Pu-i(A)

Pu-i = P(u-iA), A e B(X2).

It is well known that if u is a continuous function, then it is (B(Xi), B(X2))-measurable.

In the sequel, the following property of weakly convergent probability measures will be very useful.

Lemma 2. Suppose that Pn, n e N, and P are probability measures on (Xi, B(Xi)); the function u : Xi ^ X2 is continuous, and Pn converges weakly to P as n ^ to. Then Pnu-i also converges weakly to Pu-i as n ^ to.

The lemma is Theorem 5.1 from [1].

Lemma 3. Suppose that the set L(a, P) is linearly independent over Q, and F : H2(D) ^ H(D) is a continuous operator. Then

Pt,f(A) d=f 1meas {t g [0,T] : F (((s + iT, a)) G A} , A G B(H(D)), converges weakly to PzF-1 as T ^ to.

PROOF. The definitions of PT and PT,F imply that PT,F = PTF-1. Therefore, the continuity F

Let V > 0, md, for A e B (H2(Dy ,D)),

Pt,v(A) = ^meas {t g [0, T] : Z(s + iT, a) e A} ,

Pc,y(A) = mg (w e Q : Z(s, a, w) e A .

Lemma 4. Suppose that the set L(a, P) is linearly independe nt over Q, an d F : H 2(Dy, D) ^ H(Dy) ¿s a continuous operator. Then

PT,F,y (A) d=f ^meas {t g [0, T] : F (Z(s + iT, a)) e A} , A e B(H(Dy)), converges weakly to Pz,yF-1 as T ^ to.

Proof. Clearly, the function uy : H2(D) ^ H2(Dy, D) given bv the formula

uV(g1(s),g2(s)) = g1(s) n , g2(s) I , g1,g2 G H(D),

\ seDy /

iНе можете найти то, что вам нужно? Попробуйте сервис подбора литературы.

is continuous, and, PT,y = PTu-1. Therefore, Lemmas 1 and 2 implv that PT,y converges weakly to Pz,V as T ^ to. Since PT)F)V = PT,VF-1, to have that PT)F)V converges weakly to Pz,yF-1 as T ^to. ' ' ' ' '

Now we consider the supports of the limit measures Pz, PzF-1, Pz,y and Pz,VF-1.

Lemma 5. Suppose that the set L(a, P) is linearly independe nt over Q. Then the support of the measure Pz is the set S x H(D).

Proof. Denote bv m1H and m2H the probability Haar measures on (Q1, B(Q1)) and (Q2, B(Q2)), respectively. Then we have that mH is the product of m1H and m2H, i.e., if A = A1 x A2, where A1 e B(Q1) md A2 e B(Q2), then

mg (A) = m1H (A1 )m2g (A2). (2)

The space H2(D) is separable, therefore, B(H2(D)) = B(H(D)) x B(H(D)). Thus, it suffices to consider the measure Pz on the sets A = A1 x A2, A1, A2 e H (D). It is known [20] that the support of the measure

m1H (W1 G Q1 : Z(s, W1) G A), A e B(H(D)) (3)

is the set S. The linear independence of L(a, P) implies that of the set L(a) = {log(m+a) : m e N0}. Therefore, the case r = 1 of Theorem 11 from [6] gives that the support of the measure

m2H (w2 e Q2 : Z(s,a,w2) e A), A e B(H(D)), (4)

H(D)

Pz(A) = mH (w e Q : ((s,a,w) e A) , A e B(H2(D)),

in view of (2), we have that, for A = Ai x A2,

Pz (A) = miH (wi e Qi : ((s, wi) e Ai) m2H (w2 e Q2 : Z (s, a, w2) e A2).

Therefore, the lemma follows from remarks on supports of the measures (3) and (4), and minimality property of a support.

Lemma 6. Suppose that the set L(a, P) is linearly independent over Q, and F : H2(D) ^ H(D) is a continuous operator such that, for every open set G C H(D) the set (F-iG) n (S x H(D)) is non-empty. Then the support of the measure PzF-i is the whole of H(D).

Proof. We apply standard arguments. Let g e H(D) be an arbitrary element, and G be its

F F- i G

F-iG

S x H(D). Hence, by Lemma 5, Pz(F-iG) > 0. Therefore,

P<zF-i(G) = Pz(F-iG) > 0. Since g and G are arbitrary, this proves the lemma.

In what follows, the Mergelvan theorem on the approximation of analytic functions by polynomials will be exceptionally useful [17].

Lemma 7. Suppose that K C C is a compact subset with connected complement, and f (s) is a continuous function on K which is analytic in the interior of K. Then, for every e > 0; there exists p(s)

sup |f(s) — p(s)| < e.

Lemma 8. Suppose that the set L(a, P) is linearly independe nt over Q, an d V > 0. Then the support of Pz,y is the set Sy x H(D).

Proof. Let g be an arbitrary element of Sy x H(D), and G be its open neighborhood. The function uy defined in the proof of Lemma 4 is continuous. Therefore, by the definition of uy, the set u-iG is open and non-empty. Really, it is well known, see, for example, [8], that the

H(D)

p(s) p(s) e G

Since the polynomial p(s) is an entire function, p(s) also belongs to u-iG. Thus, the set u-iG is

S x H(D)

Pz(u-iG) > 0 Hence, Pz,y(G) = Pzu-i(G) = Pz(u-iG) > 0 Clearly, if (gi,g2) e S x H(D), then also (gi,g2) e Sy x H(D). Therefore,

mH (w e Q : ((s, a, w) e Sy x H(D)) ^ mH (w e Q : ((s, a, w) e S x H(D)) = 1.

Hence,

Pz,y (Sy x H(D)) = 1.

Lemma 9. Suppose that the set L(a, P) is linearly independent over Q. Let F : H2(Dy, D) ^ H(Dy) he a continuous operator such that, for each polynomial p = p(s), iAe set (F-1{p}) n (Sy x H(D)) is non-empty. Then the support of the measure Pz,yF-1 is the whole of H(Dy).

Proof. Let g be an arbitrary element of H(Dy), and G be its arbitrary open neighbourhood. Then, by Lemma 7, there exists a polynomial p(s) G G. Therefore, the hypotheses of the lemma imply that the set F-1G is open and contains an element of the set Sy x H(D). Thus, in virtue of Lemma 8, Pz,y(F-1G) > 0. From this, it follows that

Pc,VF-1(G) = Pzy (F-1G) > 0, and the lemma is proved because g and G are arbitrary.

Lemma 10. Suppose that the set L(a, P) is linearly independent over Q; and the operator F : H2(D) ^ H(D) satisfies the hypotheses of Theorem 6. Then the support of the measure PzF contains the closure of the set Hai)...,ar(D).

PROOF. Since F(S x H(D)) D Hai;...;ar(D), for each element g G Hai)...,ar(D), there exists an element (g1 ,g2) G S x H(D)) such that F(g1,g2) = g. If G is an arbitrary open neighborhood of g, then we have that the open set F-1G is an open neighborhood of a certain element of S x H(D). Therefore, in view of Lemma 5, Pz(F-1G) > 0. Hence,

PzF-1(G) = Pz(F-1G) > 0.

This shows that the element g lies in the support of the measure PzF-1. Since g is an arbitrary element of Hai)...,ar(D), we have that the support of PzF-1 contains the set Hai)...,ar(D). However, the support is a closed set, therefore, it contains the closure of Hai)...,ar(D).

Lemma 11. Suppose that the set L(a, P) is linearly independent over Q, and F : H2(D) ^ H(D) is a continuous operator. Then the support of PzF-1 is the closure of F(S x H(D)).

g F(S x H(D)) G

by Lemma 5, Pz(F-1G) > 0. Hence, PzF-1(G) > 0. Moreover, by Lemma 5 again,

PzF-1 (F(S x H(D))) = Pz (S x H(D)) = 1.

Therefore, the support of PzF-1 is the closure of F(S x H(D)).

" □

3. Proof of universality theorems

We will apply the equivalent of the weak convergence of probability measures in terms of continuity sets. We remind that A e B(X) is a continuity set of the probabilitv measure P on (X, B(X)) if P(dA) = 0, where dA is the boundary of A.

Lemma 12. Let Pn; n e N, and P be probability measures on (X, B(X)). Then Pn, as n ^ to, converges weakly to P if and only if, for every continuity set A of P,

lim P„(A) = P(A).

n^-rx

A proof of the lemma can be found infl], Theorem 2.1. Proof of Theorem 2. Put

Ge = ^ (gi,g2) e H (D) : sup |gi(s) — fi(s)| < e, sup |g2(s) — f2(s)| < e

I sg—i sg—

Then Ge is an open set in H2(D). Moreover,

dG£ = <j (gi,g2) e H (D) : sup |gi(s) — fi(s)| < e, sup |g2(s) — f2(s)| = e

sg—i sg—

U <J (gi,g2) e H2(D) : sup |gi(s) — fi(s)| = e, sup |g2(s) — f2(s)| < e seKi se—

^ (gi,g2) e H (D) : sup |gi(s) — fi(s)| = e, sup |g2(s) — f2(s)| = e i seKi se—

Therefore, if ei > 0 e2 > 0 and ei = e2, then dG£1 n dG£2 = 0. Hence, we have that Pz(dGe) > 0 for at most a countable set of values of e > 0. This means that the set Ge is a continuity set of Pz for all but at most countablv many e > 0. Therefore, by Lemmas 1 and 12,

lim ¿measjr e [0; T] : ((s + ir) e Ge) = Pz(Ge), t^^ 11 ~ J —

1

T ^^

or, by the definition of Ge,

lim ^measjr e [0; T] : sup |((s + ir) — fi(s)| < e,

1 L seKi

sup |((s + ir,a) — f2(s)| <4 = Pz(Ge) (5)

se— J ^

for all but at most countablv many e > 0. By Lemma 7, there exist polvnomials pi(s) and p2(s) such that

epi(s)| e

sup |fi(s) — epi(s)| <- (6)

and

sG—l ' 2

e

sup |f2(s) — p2(s)| < -. (7)

sg—2 2

In view of Lemma 5, {epi(s),p2(s)} is and element of the support of the measure Pz- Therefore, putting

G£ = j(gi,g2) e H2(D) : sup |gi(s) — epi(s)| < 2, sup |g2(s) — p2(s)| < 2) ,

i sg—i 2 sg-2 2 j

we obtain that Pz(Ge) > 0. Inequalities (6) and (7) show, that for (gi,g2) e Ge,

sup |gi(s) — fi(s)| < e

sg—i

and

sup |g2(s) — f2(s)| < e.

sg—2

Thus, we have that Ge C Ge. Hence, Pz(Ge) ^ Pz(Ge) > 0. This together with (5) proves the theorem.

Proof of Theorem 4. Define the set

GM = g e H(D) : sup |g(s) - f (s)| < e . I seK J

Then we have that G1;£ is a continuity set of the measure P^F-1 for all but at most countablv many e > 0.Hence, in view of Lemmas 3 and 12,

lim —meas {r G [0; T] : F (Z(s + ît, a)) G GM}

= lim — measit g [0; T] : sup |F(Z(s + ît), Z(s + ît, a)) - f (s)| < e) T L seK J

=PÇF-1(GM) (8)

for all but at most countablv many e > 0. By Lemma 7, there exists a polvnomial p(s) such that

iНе можете найти то, что вам нужно? Попробуйте сервис подбора литературы.

Define

£

sup |f (s) - p(s)| < -. (9)

seK 2

GM = <j g g H(D) : sup |g(s) - f (s)|

seK -

The polynomial p(s), by Lemma 6, is an element of the support of the measure P^F 1- Hence, P<Z(G 1,e) > 0. Obviously, for g e G 1)£, by (9),

sup |g(s) - f (s)| < e.

seK

Therefore, G1)£ C G1)£, P^F-1(G1;£) ^ P^F-1(G1;£) > 0, and the theorem follows from (8).

' ' " ' " ' □

proof of Theorem 5. We follow the proof of Theorem 4, and use Lemma 4 in place of Lemma 3, and Lemma 9 in place of Lemma 6.

Proof of Theorem 6. The case r = 1. By Lemma 7, there exists a polvnomial p(s) such that

£

sup |f (s) - p(s)| <-. (10)

seK 4

By hypotheses of the theorem, f (s) = a1 on K. Therefore, in view of (10), p(s) = a1 on K as well if e is small enough. Thus, we can define a continuous branch of log(p(s) — a1) which will be an analytic function in the interior of K. Using Lemma 7 once more, we find a polynomial p1 (s) such that

sup |p(s) - fl! - epi(s))| <e. (11)

seK 4

Now we put f (s) = Then f1(s) e H(D) and f1(s) = a^. Therefore, by Lemma 10, f1(s)

is an element of the support of the measure P^F-1. Define

Gm = { g e H (D) : sup |g(s) - f1(s)| <e

L seK 2

Then £i,£ is an open neighborhood of /^s), thus, PzF 1(Gi,e) > 0. Now consider the set

Gi,e = g e H (D) : sup |g(s) - / (s)| <e . I seK J

Similarly as in the proof of the above theorems, we observe that G1,e is an continuity set of the measure Pc F-1 for all but at most countablv many e > 0. Therefore, taking into account Lemmas 3 and 12, we have that

lim — meas {r e [0; T] : F (Z(s + ir, a)) e Gi,e )

T—^ T L )

= lim ¿meaJr e [0; T] : sup |F(Z(s + ir, a)) - /(s)| <4 = PzF-1 (Gi,e). (12)

T—<x T I _ J _

Clearly, by (10) and (11),

e

sup |/(s) - /i(s)| < -.

seK 2

Therefore, if g e G^e, then g e G^e, i-e., Gi,e C G^e- Since PzF-i(Gi,e) > 0, we have that PzF-i(G/i,e) > 0. This inequality together with (12) proves the theorem in the case r = 1. Now let r ^ 2. Define

G2,e = g e H(D) : sup |g(s) - /(s)| < e . I J

Since /(s) e Hai,...,ar(D), we have by Lemma 10, that /(s) is an element of the support of PzF-i. Moreover, G2,e is an open neighborhood of /(s). Therefore,

P<zF-i (G2,e) > 0. (13)

On the other hand, G2,e is a continuity set of the measure PzF-i for all but at most countablv many e > 0. Therefore, in view of Lemmas 3 and 12, and (12)

lim ^measj r e [0; T] : sup |F (Z (s + ir, a)) - / (s)| < el T—<^ T I seK " J

= lim ^meas {r e [0; T] : F (Z(s + ir, a)) e G2,e} = PzF-i(G2,e) > 0.

T—t —

Proof of Theorem 7. We repeat the proof of the case r ^ 2 of Theorem 6, and, in place of Lemma 10, we apply Lemma 11.

4. Conclusions

Z(s) Z(s, a)

transcendental or rational parameter a are universal in the Voronin sense, i.e., their shifts Z(s + ir) and Z(s + ir, a) r e R, approximate functions from wide classes. H. Mishou obtained a joint universality theorem for Z(s) and Z(s, a). He proved that the set of shifts (Z(s + ir), Z(s + ir, a)) a

In the paper, it is observed that the set of the above shifts has a positive density for all but at most countablv many values of e > 0, where e is accuracy of approximation.

Also, it is obtained that composite functions F(Z(s),Z(s, a)) for some classes of operators F

H(D)

shifts F(Z(s + ¿t), Z(s + ¿t, a)) approximating a given analytic function with accuracy e > 0 has a positive density for all but at most countablv many values of e.

REFERENCES

1. Billingslev P., 1968, "Convergence of Probability Measures" , New York: Willey.

2. Javtokas A., Laurincikas A., 2006, "Universality of the periodic Hurwitz zeta-function" , Integr. Transf. Spec. Fund. Vol. 17. P. 711-722.

3. Kacinskaite R., Laurincikas A., 2011, "The joint distribution of periodic zeta-functions" , Studia Sci. Math. Hung. Vol. 18. P. 257-279.

4. Laurincikas A., 2007, "Voronin-tvpe theorem for periodic Hurwitz zeta-functions" , Sb. Math. Vol. 198, No. 1-2. P. 231-242.

5. Laurincikas A., 2008, "Joint universality for periodic Hurwitz zeta-functions" , Izv. Math. Vol. 72, No. 1-2. P. 741-760.

6. Laurincikas A., 2008, "The joint universality of Hurwitz zeta-functions" , Siauliai Math. Semin. Vol. 3(11). P. 169-187.

7. Laurincikas A., 2010, "Joint universality of zeta-functions with periodic coefficients" , Izv. Math. Vol. 74. P. 515-539.

8. Laurincikas A., 2012, "Universality of composite functions" , in: Functions in Number Theory and Their Probabilistic Aspects, K. Matsumoto et al (Eds), RIMS Kokvuroku Bessatsu. Vol. B34. P. 191-204.

9. Laurincikas A., 2012, "On joint universality of the Riemann zeta-function and Hurwitz zeta-functions" , J. Number Theory. Vol. 132. P." 2842-2853.

10. Laurincikas A., 2016, "Universality theorems for zeta-functions with periodic coefficients" , Sib. Math. J. Vol. 57, No. 2. P. 330-339.

11. Laurincikas A., Garunkstis R., 2002, "The Lerch Zeta-Function" , Dordrecht: Kluwer.

12. Laurincikas A., Matsumoto K., 2001, "The universality of zeta-functions attached to certain cusp forms" , Acta Arith. Vol. 98. P. 345-359.

13. Laurincikas A., Meska L., 2014, "Improvement of the universality inequality" , Math. Notes. Vol. 96, No. 5-6. P. 971-976.

14. Laurincikas A., Meska L., 2016, "On the modification of the universality of the Hurwitz zeta-function" , Nonlinear Analysis: Modelling and Control. Vol. 21, No. 4. P. 564-576.

15. Laurincikas A., Siauciunas D., 2006, "Remarks on the universality of the periodic zeta-function" , Math. Notes. Vol. 80, No. 3-4. P. 532-538.

16. Matsumoto K., 2015, "A survey on the theory of universality for zeta and L-functions" , in: Series on Number Theory and Its Applications. Number Theory: Plowing and Starring Through High Wave Forms, Proc. of the 7th China-Japan Seminar, Fukuoka, Japan, 2013, M. Kaneko ed al (Eds). Vol. 11. P. 95-144.

17. Mergelvan S.N., 1952, "Uniform approximation to functions of a complex variable" , Uspekhi Mat. Nauk Vol. 7. P. 31 122. (In Russian).

18. Meska L., 2014, "A modification of the universality inequality" , Siauliai Math. Semin. Vol. 9(17). P. 71-81.

19. Mishou H., 2007, "The joint value distribution of the Riemann zeta-function and Hurwitz zeta-functions" , Lith. Math. J. Vol. 47. P. 32-47.

20. Steuding J., 2007, "Value-Distribution of L-functions" , Lecture Notes in Math. 1877, Berlin: Springer.

21. Voronin S.M., 1975, "A theorem on the "universality"of the Riemann zeta-function" , Math. USSR Izv. Vol. 9. P. 443-453.

СПИСОК ЦИТИРОВАННОЙ ЛИТЕРАТУРЫ

1. Billingslev P. Convergence of Probability Measures. New York: Willev, 1968.

2. Javtokas A., Laurincikas A. Universality of the periodic Hurwitz zeta-function // Integr. Transf. Spec. Funct. 2006. Vol. 17. P. 711-722.

3. Kacinskaite R., Laurincikas A. The joint distribution of periodic zeta-functions // Studia Sci. Math. Hung. 2011. Vol. 18. P. 257-279.

4. Лауринчикас А.П. Аналог теоремы Воронина для периодических дзета-функций Гурвица // Матем. Сб. 2007. Т. 198, №. 2. С. 91-102.

5. Лауринчикас А. Совместная универсальность периодических дзета-функций Гурвица // Изв. РАН. Сер. матем. 2008. Т. 72, №. 4. С. 121-140.

6. Laurincikas A. The joint universality of Hurwitz zeta-functions // Siauliai Math. Semin. 2008. Vol. 3(11). P. 169-187.

7. Лауринчикас А. Совместная универсальность дзета- функций с периодическими коэффициентами // Изв. РАН. Сер. матем. 2010. Т. 74, №. 3. С. 79-102.

8. Laurincikas A. Universality of composite functions // Functions in Number Theory and Their Probabilistic Aspects, K. Matsumoto et al (Eds), RIMS Kokvuroku Bessatsu. 2012. Vol. B34. P. 191-204.

9. Laurincikas A. On joint universality of the Riemann zeta-function and Hurwitz zeta-functions // J. Number Theory. 2012. Vol. 132. P. 2842-2853.

10. Лауринчикас А. Расширение универсальности дзета функций с периодическими коэффициентами // Сиб. матем. ж. 2016. Т. 57, №. 2. С. 420-431.

11. Laurincikas A., Garunkstis R. The Lerch Zeta-Function. Dordrecht: Kluwer, 2002.

12. Laurincikas A., Matsumoto K. The universality of zeta-functions attached to certain cusp forms // Acta Arith. 2001. Vol. 98. P. 345-359.

13. Лауринчикас А., Мешка Л. Уточнение неравенства универсальности // Матем. заметки. 2014. Т. 96, №. 6. С. 905-910.

14. Laurincikas A., Meska L. On the modification of the universality of the Hurwitz zeta-function // Nonlinear Analysis: Modelling and Control. 2016. Vol. 21, No. 4. P. 564-576.

15. Лауринчикас А.П., Шяучюнас Д. Замечания об универсальности периодической дзета-функции // Матем. заметки. 2006. Т. 80, №. 4. С. 561-568.

16. Matsumoto К. A survey on the theory of universality for zeta and L-functions // in: Series on Number Theory and Its Applications. Number Theory: Plowing and Starring Through High Wave Forms, Proc. of the 7th China-Japan Seminar, Fukuoka, Japan, 2013. M. Kaneko ed al (Eds). 2015. Vol. 11. P. 95-144.

17. Мергелян C.H. Равномерные приближения функций комплексного переменного // УМН. 1952. Т. 7, №. 2. С. 31-122

18. Meska L. A modification of the universality inequality // Siauliai Math. Semin. 2014. Vol. 9(17). P. 71-81.

19. Mishou H. The joint value distribution of the Riemann zeta-function and Hurwitz zeta-functions // Lith. Math. J. 2007. Vol. 47. P. 32-47.

20. Steuding J. Value-Distribution of L-functions, Lecture Notes in Math. 1877. Berlin: Springer, 2007.

21. Воронин С. M. Теорема об "универсальности" дзета-функции Римана // Изв. АН СССР. Сер. матем. 1975. Т. 39. С. 475-486.

Vilnius University. Получено 27.06.2016 г. Принято в печать 12.09.2016 г.

i Надоели баннеры? Вы всегда можете отключить рекламу.