APPROXIMATION BY T MEANS WITH RESPECT TO

VILENKIN SYSTEM IN LEBESGUE SPACES AND

arXiv:2405.19350v1 [math.GM] 21 May 2024

LIPSCHITZ CLASSES

N. ANAKIDZE, N. ARESHIDZE, L.-E. PERSSON AND G. TEPHNADZE

Abstract. In this paper we present and prove some new results con-
cerning approximation properties of T means with respect to the Vilenkin
system in Lebesgue spaces and Lipschitz classes for any 1 ≤ p < ∞. As
applications, we obtain extension of some known approximation inequal-
ities.

2010 Mathematics Subject Classification: 42C10.

Key words and phrases: Vilenkin group, Vilenkin system, T means, Nör-
lund means, Fejér means, approximation, Lebesgue spaces, Lipschitz classes.

1. Introduction
Concerning some definitions and notations used in this introduction we
refer to Section 2.
It is well-known (see e.g. [17], [29] and [40]) that, for any 1 ≤ p ≤ ∞ and
f ∈ Lp (Gm ), there exists an absolute constant Cp , depending only on p such
that

kσn f kp ≤ Cp kf kp .
Moreover, (see e.g. [29]) if 1 ≤ p ≤ ∞, MN ≤ n < MN +1 , f ∈ Lp (Gm ) and
n ∈ N, then

N
X Ms
(1) kσn f − f kp ≤ 2R5 ωp (1/Ms , f ) ,
MN
s=0

where R := supk∈N mk and ωp (δ, f ) is the modulus of continuity of Lp func-

tions, 1 ≤ p < ∞ functions defined by

ωp (δ, f ) = sup kf (x − t) − f (x)kp , δ > 0.

|t|<δ

It follows that if f ∈ Lip (α, p) , i.e.,

Lip(α, p) := {f ∈ Lp : ωp (δ, f ) = O(δα ) as δ → 0},

The research was supported by Shota Rustaveli National Science Foundation (SRNSF)
grant no. FR-24-698.
1
2 N. ANAKIDZE, N. ARESHIDZE, L.-E. PERSSON AND G. TEPHNADZE

then 
 O (1/MN ) , if α > 1,
kσn f − f kp = O (N/MN ) , if α = 1,

O (1/Mnα ) , if α < 1.
Moreover, (see e.g. [29]) if 1 ≤ p < ∞, f ∈ Lp (Gm ) and

kσMn f − f kp = o (1/Mn ) , as n → ∞,
then f is a constant function.
For the maximal operators of Vilenkin-Fejer means σ ∗ , defined by
σ ∗ f = sup |σn f |
n∈N

the weak-(1, 1) type inequality

kσ ∗ f kweak−L1 ≤ C kf k1 , f ∈ L1 (Gm )
can be found in Schipp [30] for Walsh series and in Pál, Simon [28] and Weisz
[38] for bounded Vilenkin series. Boundedness of the maximal operators of
Vilenkin-Féjer means of the one- and two-dimensional cases can be found
in Fridli [12], Gát [14], Goginava [16], Nagy and Tephnadze [24, 25, 26, 27],
Simon [32, 33] and Weisz [39].
Convergence and summability of Nörlund means with respect to Vilenkin
systems were studied by Areshidze and Tephnadze [2], Blahota and Nagy [5],
Blahota, Persson and Tephnadze [9] (see also [3, 4, 7, 8]), Fridli, Manchanda
and Siddiqi [13], Goginava [15], Nagy [21, 22, 23] (see also [10] and [11]) and
Memic [18].
Móricz and Siddiqi [20] investigated the approximation properties of some
special Nörlund means of Walsh-Fourier series of Lp functions in norm. In
particular, they proved that if f ∈ Lp (G), 1 ≤ p ≤ ∞, n = 2j + k, 1 ≤ k ≤
2j (n ∈ N+ ) and (qk , k ∈ N) is a sequence of non-negative numbers, such
that
n−1
nγ−1 X γ
qk = O(1), for some 1 < γ ≤ 2,
Qγn
k=0
then there exists an absolute constant Cp , depending only on p such that

j−1    
Cp X i 1 1
ktn f − f kp ≤ 2 qn−2i ωp , f + Cp ω p ,f ,
Qn 2i 2j
i=0
when the sequence (qk , k ∈ N) is non-decreasing, while

j−1    
Cp X 1 1
ktn f − f kp ≤ (Qn−2i +1 − Qn−2i+1 +1 ) ωp , f + C ω
p p , f ,
Qn 2i 2j
i=0

when the sequence (qk , k ∈ N) is non-increasing.

 APPROXIMATION BY T MEANS 3

Tutberidze [34] (see also [29]) proved that if Tn are T means generated by
either a non-increasing sequence {qk , k ∈ N} or a non-decreasing sequence
{qk , k ∈ N} satisfying the condition
 
q0 1
=O , as k → ∞,
Qk k
then there exists an absolute constant C, such that


kT ∗ f kweak−L1 ≤ C kf k1 , f ∈ L1 (Gm )
holds. From these results follows that if f ∈ Lp (Gm ), where 1 ≤ p < ∞
and either the sequence {qk , k ∈ N} is non-increasing, or {qk , k ∈ N} is a
sequence of non-decreasing numbers, such that the condition
 
qn−1 1
(2) =O , as n → ∞,
Qn n
is fulfilled, then
lim kTn f − f kp → 0, as n → ∞.
n→∞
For the Walsh system in [19] Móricz and Rhoades proved that if f ∈ Lp ,
where 1 ≤ p < ∞, and Tn are regular T means generated by a non-increasing
sequence {qk , k ∈ N}, then, for any 2N ≤ n < 2N +1 , we have the following
approximation inequality:

N −1
Cp X s

(3) kTn f − f kp ≤ 2 q2s ωp (1/2s , f ) + Cp ωp 1/2N , f .
Qn
s=0
In the case when the sequence {qk , k ∈ N} is non-decreasing and satisfying
the condition
 
qk−1 1
(4) =O , as k → ∞,
Qk k
then the following inequality holds:

N
X −1



(5) kTn f − f kp ≤ Cp 2j−N ωp 1/2j , f + Cp ωp 1/2N , f .
j=0

In this paper we use a new approach and generalize the inequalities in (3)
and (5) for T means with respect to the Vilenkin system (see Theorems 1
and 2). We also prove a new inequality for the subsequences {TMn } means
when the sequence {qk , k ∈ N} is non-decreasing (see Theorem 3).
The main results are presented in Section 3 together with 4 Corollar-
ies, which also generalize known results (see [19]). The detailed proofs are
presented in Section 4. In order not do disturb our presentations in these
Sections we have reserved Section 2 for some Preliminaries.
4 N. ANAKIDZE, N. ARESHIDZE, L.-E. PERSSON AND G. TEPHNADZE

2. Preliminaries
Let N+ denote the set of the positive integers, N := N+ ∪ {0}. Let m =:
(m0 , m1 , ...) be a sequence of positive integers not less than 2. Denote by
Zmk := {0, 1, ..., mk − 1}
the additive group of integers modulo mk . Define the group Gm as the com-
plete direct product of the group Zmk with the product of the discrete topolo-
gies of Zmk ’s.
The direct product µ of the measures
µk ({j}) := 1/mk (j ∈ Zmk )
is the Haar measure on Gm with µ (Gm ) = 1.
If supk∈N mk < +∞, then we call Gm a bounded Vilenkin group. If the
sequence {mk }k≥0 is unbounded, then Gm is said to be unbounded Vilenkin
group. In this paper we consider only bounded Vilenkin groups.
The elements of Gm are represented by the sequences

x := (x0 , x1 , . . . , xk , . . . ) ( xk ∈ Zmk ) .
It is easy to give a base for the neighborhood of Gm , namely

I0 (x) := Gm , In (x) := {y ∈ Gm | y0 = x0 , . . . , yn−1 = xn−1 } (x ∈ Gm , n ∈ N).

For simplicity we also define In := In (0).
Next, we define a generalized number system based on m in the following
way:
M0 =: 1, Mk+1 =: mk Mk (k ∈ N)
Then every n ∈ N can be uniquely expressed as
∞
X
n= nj Mj , where nj ∈ Zmj (j ∈ N)
k=0
and only a finite number of nj ‘s differ from zero. Let

|n| =: max{j ∈ N, nj 6= 0}.

Moreover, Vilenkin (see [35, 36, 37]) investigated the group Gm and intro-
duced the Vilenkin system {ψj }∞j=0 as
∞
Y
ψn (x) := rknk (x) (n ∈ N) .
k=0

where rk (x) are the generalized Rademacher functions defined by

rk (x) := exp(2πixk /mk ), (k ∈ N) .
These systems include as a special case the Walsh system when mk = 2
for any k ∈ N.
 APPROXIMATION BY T MEANS 5

The norms (or quasi-norms) kf kp , 0 < p < ∞, of the Lebesgue spaces

Lp (Gm ) are defined by
Z
p
kf kp := |f |p dµ.
Gm

The Vilenkin system is orthonormal and complete in L2 (Gm ) (see e.g. [1]
and [31]).
If f ∈ L1 (Gm ), we can define the Fourier coefficients, the partial sums
of the Fourier series, the Fejér means, the Dirichlet and Fejér kernels with
respect to the Vilenkin system in the usual manner:

Z
fb(k) : = f ψ k dµ, (k ∈ N) ,
Gm
n−1
X
Sn f : = fb(k) ψk , (n ∈ N+ , S0 f := 0) ,
k=0
n
1X
σn f : = Sk f, (n ∈ N+ ) .
n
k=1
n−1
X
Dn : = ψk , (n ∈ N+ ) .
k=0
n
1X
Kn : = Dk , (n ∈ N+ ) .
n
k=1

Recall that (see e.g. [1] and [29]),


Mn , if x ∈ In ,
(6) DMn (x) =
0, if x ∈
/ In ,

(7) DMn −j (x) = DMn (x) − ψ Mn −1 (−x)Dj (−x)

= DMn (x) − ψMn −1 (x)Dj (x), 0 ≤ j < Mn .

|n|
X
2
(8) n |Kn | ≤ 2R Ml |KMl | ,
l=0

and

Z Z
(9) Kn (x)dµ(x) = 1, sup |Kn (x)| dµ(x) ≤ R5 .
Gm n∈N Gm

where R := supk∈N mk . Moreover, if n > t, t, n ∈ N, then

6 N. ANAKIDZE, N. ARESHIDZE, L.-E. PERSSON AND G. TEPHNADZE

 Mt
 1−rt (x) , x ∈ It \It+1 , x − xt et ∈ In ,
(10) KMn (x) = Mn +1
 2 , x ∈ In ,
0, otherwise.

The n-th Nörlund mean tn and T mean Tn of f ∈ L1 (Gm ) are defined by

n
1 X
tn f := qn−k Sk f
Qn
k=1

and
n−1
1 X
Tn f := qk Sk f,
Qn
k=0
where
n−1
X
Qn := qk .
k=0

Here {qk , k ≥ 0} is a sequence of nonnegative numbers, where q0 > 0 and

(11) lim Qn = ∞.
n→∞

Then, a T mean generated by {qk , k ≥ 0} is regular if and only if the

condition (11) is fulfilled (see [29]).
It is evident that
Z
Tn f (x) = f (t) Fn (x − y) dµ (y)
Gm

where
n−1
1 X
(12) Fn := qk Dk ,
Qn
k=0

which are called the kernels of the T means.

By applying Abel transformation we get the following two useful identities:

n−1
X n−2
X
(13) Qn := qk · 1 = (qk − qk+1 )k + qn−1 (n − 1)
k=0 k=0

and

n−2
!
1 X
(14) Tn f = (qk − qk+1 )kσk f + qn−1 (n − 1)σn−1 f .
Qn
k=0
 APPROXIMATION BY T MEANS 7

3. The Main Results

Our first main result reads:
Theorem 1. Let f ∈ Lp (Gm ), where 1 ≤ p < ∞ and Tn are T means
generated by a non-increasing sequence {qk , k ∈ N}. Then, for any n, N ∈ N,
MN ≤ n < MN +1 , we have the following inequality:

N −1
6R6 X
(15) kTn f − f kp ≤ Mj qMj ωp (1/Mj , f ) + 4R6 ωp (1/MN , f ) .
Qn
j=0

Next we state and prove a similar inequality for non-decreasing sequences

but under some restrictions.
Theorem 2. Let f ∈ Lp (Gm ), where 1 ≤ p < ∞ and Tn are regular T
means generated by a non-decreasing sequence {qk , k ∈ N}. Then, for any
n, N ∈ N, MN ≤ n < MN +1 , we have the following inequality:

(16)
N −1
6R6 qn−1 X 4R6 qn−1 MN
kTn f − f kp ≤ Mj ωp (1/Mj , f ) + ωp (1/MN , f ) .
Qn Qn
j=0

If, in addition, the sequence {qk , k ∈ N} satisfies the condition (2), then the
inequality

N
X Mj

(17) kTn f − f kp ≤ Cp ωp 1/2j , f
MN
j=0

holds for some constant Cp only depending on p.

Finally, we state and prove the third main result for the non-decreasing
sequences, where we prove a more precise result then that in (17) and without
the restriction (2) but only for subsequences.
Theorem 3. Let f ∈ Lp (Gm ), where 1 ≤ p < ∞ and Tk are regular T means
generated by a non-decreasing sequence {qk , k ∈ N}. Then, for any n ∈ N,
the following inequality holds:

(18) kTMn f − f kp
n−1
X
2 Mj
≤ R ωp (1/Mj , f )
Mn
j=0
n−1
2R4 X (n − j)(qMn −Mj )Mj
+ ωp (1/Mj , f )
q0 Mn
j=0


+ R2 + 2 ωp (1/Mn , f ) .
8 N. ANAKIDZE, N. ARESHIDZE, L.-E. PERSSON AND G. TEPHNADZE

We also point out the following generalizations of some results in [19] (in
this paper only the Walsh system was considered):
Corollary 1. Let {qk , k ≥ 0} be a sequence of non-negative and non-
increasing numbers, while in case when the sequence is non-decreasing it
is assumed that also the condition (2) is satisfied. If f ∈ Lip(α, p) for some
α > 0 and 1 ≤ p < ∞, then

 O(n−α ), if 0 < α < 1,
kTn f − f kp = O(n−1 log n), if α = 1,

O(n−1 ), if α > 1,

Corollary 2. Let {qk , k ≥ 0} be a sequence of non-negative and non-

increasing numbers such that

qk ≍ k−β for some 0<β≤1

is satisfied.
If f ∈ Lip(α, p) for some α > 0 and 1 ≤ p < ∞, then



 O(n−α ), if α + β < 1,

O(n−(1−β) log n + n−α ), if α + β = 1,
kTn f − f kp =

 O(n−(1−β) ), if α + β > 1, β > 1,

O((log n)−1 ), if β = 1.

Corollary 3. Let {qk , k ≥ 0} be a sequence of non-negative and non-

increasing numbers such that

qk ≍ (log k)−β for some β>0

is satisfied.
If f ∈ Lip(α, p) for some α > 0 and 1 ≤ p < ∞, then



 O(n−α ), if 0 < α < 1, β > 0,

O(n−1 log n), if α = 1, 0 < β < 1,
kTn f − f kp =

 O(n−1 log n log log n), if α = β = 1,

O(n−1 (log n)β ), if α > 1, β > 0.

Corollary 4. Let f ∈ Lp (Gm ), where 1 ≤ p < ∞ and {qk , k ≥ 0} is a

sequence of non-negative and non-increasing numbers, while in case when
the sequence is non-decreasing it is also assumed that the condition (2) is
satisfied. Then,

lim kTn f − f kp → 0, as n → ∞.
n→∞
 APPROXIMATION BY T MEANS 9

4. Proofs
Proof of Theorem 1. Let MN ≤ n < MN +1 . Since Tn are regular T means
generated by the sequence of non-increasing numbers {qk : k ∈ N}, we can
combine (13) and (14) and conclude that

(19) kTn f − f kp
 
n−2
X
1 
≤ (qj − qj+1 )jkσj f − f kp + qn−1 (n − 1)kσn−1 f − f kp 
Qn
j=0
:= I + II.

Moreover,

MN −1
1 X
(20) I = (qj − qj+1 ) jkσj f − f kp
Qn
j=1
n−1
X
1
+ (qj − qj+1 ) jkσj f − f kp
Qn
j=MN
:= I1 + I2 .

Now we estimate all terms separately. By applying estimate (1) for I1 we

obtain that

N −1Mk+1 −1 k
2R5 X X X Ms
(21) I1 ≤ (qj − qj+1 ) j ωp (1/Ms , f )
Qn s=0
Mk
k=0 j=Mk
N −1 Mk+1 −1 k
2R6 X X X Ms
≤ Mk (qj − qj+1 ) ωp (1/Ms , f )
Qn Mk
k=0 j=Mk s=0

6N
X −1  X
k
2R
≤ qMk − qMk+1 Ms ωp (1/Ms , f )
Qn
k=0 s=0
N −1 N −1  
2R6 X X
≤ Ms ωp (1/Ms , f ) qMk − qMk+1
Qn s=0 k=s
6N
X −1
2R
≤ Ms qMs ωp (1/Ms , f ) .
Qn
s=0

Moreover,
10 N. ANAKIDZE, N. ARESHIDZE, L.-E. PERSSON AND G. TEPHNADZE

n−1 N
2R5 X X Ms
(22) I2 ≤ (qj − qj+1 ) j ωp (1/Ms , f )
Qn MN
j=MN s=0
n−1
X N
X
2R6 MN Ms
≤ (qj − qj+1 ) ωp (1/Ms , f )
Qn MN
j=MN s=0
N
2R6 q MN
X
≤ Ms ωp (1/Ms , f )
Qn
s=0
N
2R6 X
≤ Ms qMs ωp (1/Ms , f )
Qn s=0
N −1
2R6 X
≤ Ms qMs ωp (1/Ms , f ) + 2R6 ωp (1/Ms , f ) .
Qn
s=0
For II we have that
N
2R5 MN +1 qn−1 X Ms
(23) II ≤ ωp (1/Ms , f )
Qn s=0
MN
N −1
2R6 X
≤ Ms qMs ωp (1/Ms , f ) + 2R6 ωp (1/MN , f ) .
Qn
s=0
The proof of (15) is complete by just combining (19)-(23).

Proof of Theorem 2. Let MN ≤ n < MN +1 . Since Tn are regular T means,
generated by a sequence of non-decreasing numbers {qk : k ∈ N}, by com-
bining (13) and (14), we find that
(24) kTn f − f kp
 
n−1
X
1 
≤ (qj+1 − qj ) jkσj f − f kp + qn−1 (n − 1)kσn f − f kp 
Qn
j=1
:= I + II.
Furthermore,
MN −1
1 X
(25) I = (qj+1 − qj ) jkσj f − f kp
Qn
j=1
n−1
X
1
+ (qj+1 − qj ) jkσj f − f kp
Qn
j=MN
= I1 + I2 .
Analogously to (21) we get that
 APPROXIMATION BY T MEANS 11

N −1 k
2R6 X
X
(26) I1 ≤ qMk+1 − qMk Ms ωp (1/Ms , f )
Qn
k=0 s=0
N −1 N −1  
2R6 X X
≤ Ms ωp (1/Ms , f ) qMk+1 − qMk
Qn
s=0 k=s
N −1
2R6 X
= Ms ωp (1/Ms , f ) (qMN − qMs )
Qn
s=0
N
X −1
2R6 q MN
≤ Ms ωp (1/Ms , f )
Qn
s=0
N
X −1
2R6 qn−1
≤ Ms ωp (1/Ms , f ) .
Qn s=0

In the similar way as in (22) we find that

n−1 N
2R5 X X Ms
(27) I2 ≤ (qj+1 − qj ) j ωp (1/Ms , f )
Qn MN
j=1 s=0

XN
2R5 Ms
= ((n − 1)qn−1 − Qn ) ωp (1/Ms , f )
Qn s=0
MN
N
2R5 MN +1 qn−1 X
≤ Ms ωp (1/Ms , f )
Qn MN
s=0
N
X
2R6 qn−1
≤ Ms ωp (1/Ms , f )
Qn
s=0
XN
2R6 qn−1 2R6 qn−1 MN
≤ Ms ωp (1/Ms , f ) + (1/MN , f ) .
Qn s=0
Qn
For II we have that
N
2R5 qn−1 MN +1 X Ms
(28) II ≤ ωp (1/Ms , f )
Qn s=0
MN
N
2R6 qn−1 X
≤ Ms ωp (1/Ms , f )
Qn
s=0
N
X −1
2R6 qn−1 2R6 qn−1 MN
= Ms ωp (1/Ms , f ) + (1/MN , f ) .
Qn s=0
Qn

By combining (24)-(28) we find that (16) holds. Moreover, by using con-

dition (2) we obtain the estimate (17) so the proof is complete. 
12 N. ANAKIDZE, N. ARESHIDZE, L.-E. PERSSON AND G. TEPHNADZE

Proof of Theorem 3. According to (7) we find that

Mn −1
1 X


TMn f = DMn ∗f − qk ψMn −1 Dk ∗ f .
QM n
k=0

Hence, by using Abel transformation we get that

TMn f = DMn ∗ f
Mn −2
1 X

− qMn−j − qMn −j−1 j((ψMn −1 Kj ) ∗ f )
QM n
j=0
1
− q (Mn − 1)(ψMn −1 K Mn −1 ∗ f )
QMn Mn −1
= DMn ∗ f
Mn −2
1 X

− qMn−j − qMn −j−1 j((ψMn −1 Kj ) ∗ f )
QM n
j=0
1
− q Mn (ψMn −1 K Mn ∗ f )
QMn Mn −1
qMn −1
+ (ψ D Mn ∗ f )
QMn Mn −1
so that

(29) TMn f (x) − f (x)

Z
= (f (x − t) − f (x))DMn (t)dt
Gm
Mn −2 Z
1 X

− qMn −j − qMn −j−1 j (f (x − t) − f (x)) ψMn −1 (t)K j (t)dt
QM n Gm
j=0
Z
1
− qMn −1 Mn (f (x − t) − f (x))ψMn −1 (t)K Mn (t)dt
QM n Gm
Z
qMn −1
+ (f (x − t) − f (x))ψMn −1 (t)D Mn (t)dt
Q M n Gm
=: I + II + III + IV.
By combining generalized Minkowski’s inequality and (6) we find that
Z
(30) kIkp ≤ kf (x − t) − f (x))kp DMn (t)dt ≤ ωp (1/Mn , f ) .
In

and
Z
(31) kIV kp ≤ kf (x − t) − f (x))kp DMn (t)dt ≤ ωp (1/Mn , f ) .
In
 APPROXIMATION BY T MEANS 13

Moreover, since
Mn qMn −1 ≤ QMn , for any n ∈ N,
we can use (10) and generalized Minkowski’s inequality to find that
Z
(32) kIIIkp ≤ kf (x − t) − f (x)kp K Mn (t) dµ(t)
Z Gm
= kf (x − t) − f (x)kp K Mn (t) dµ(t)
In
n−1
X mXs −1 Z

+ kf (x − t) − f (x)kp K Mn (t) dµ(t)

s=0 ns =1 In (ns es )
Z
Mn + 1
≤ kf (x − t) − f (x)kp dµ(t)
In 2
n−1
X m
Xs −1 Z

+ Ms+1 kf (x − t) − f (x)kp dµ(t)

s=0 ns =1 In (ns es )
Z
Mn + 1
≤ ωp (1/Mn , f ) dµ(t)
In 2
n−1
X X Z
ms −1
+ Ms+1 ωp (1/Ms , f ) dµ(t)
s=0 ns =1 In (ns es )
n−1
X Ms
≤ ωp (1/Mn , f ) + R2 ωp (1/Ms , f )
s=0
Mn
n
X Ms
≤ R2 ωp (1/Ms , f ) .
Mn
s=0

From this inequality and the estimates in (32) it follows also that

Z
Mn kf (x − t) − f (x)kp |K Mn (t)| dµ(t)
Gm
n
X
≤ R2 Ms ωp (1/Ms , f ) .
s=0

Let Mk ≤ j < Mk+1 By applying (8) and the last estimate we find that

Z
j kf (x − t) − f (x)kp |K j (t)|dµ(t)
Gm
k X
X l
4
≤ 2R Ms ωp (1/Ms , f ) .
l=0 s=0
14 N. ANAKIDZE, N. ARESHIDZE, L.-E. PERSSON AND G. TEPHNADZE

Hence, by also using (8) we obtain that

(33) kIIkp

Mn −1 Z
1 X

≤ qMn −j − qMn−j−1 j kf (x − t) − f (x)kp |K j (t)|dµ(t)
QM n Gm
j=0
n−1Mk+1 −1 Z
1 X X

≤ qMn −j − qMn −j−1 j kf (x − t) − f (x)kp |K j (t)|dµ(t)
QM n Gm
k=0 j=Mk
n−1Mk+1 −1 k X
l
2R4 X X
X
≤ qMn −j − qMn −j−1 Ms ωp (1/Ms , f )
QM n
k=0 j=Mk l=0 s=0

2R4 X  XX
n−1 k l
≤ qMn −Mk − qMn −Mk+1 Ms ωp (1/Ms , f )
QM n s=0
k=0 l=0

2R4 X X  X
n−1 n−1 l
≤ qMn −Mk − qMn −Mk+1 Ms ωp (1/Ms , f )
QM n
l=0 k=l s=0
n−1 l
2R4 X X
≤ qMn−Ml Ms ωp (1/Ms , f )
QM n s=0
l=0
n−1 n−1
2R X
4 X
≤ Ms ωp (1/Ms , f ) qMn −Ml
QMn s=0
l=s
n−1
2R4 X
≤ Ms ωp (1/Ms , f ) qMn −Ms (n − s)
QM n
s=0
n−1
X (n − s)Ms qMn −Ms
≤ 2R4 ωp (1/Ms , f ) .
s=0
M n q 0

Finally, by combining (30)-(33) we obtain (18) so the proof is complete.



References
[1] G. N. Agaev, N. Ya. Vilenkin, G. M. Dzhafarly and A. I. Rubinshtein, Multiplicative
systems of functions and harmonic analysis on zero-dimensional groups, Baku, Ehim,
1981 (in Russian).
[2] N. Areshidze and G. Tephnadze, Approximation by Nörlund means with respect to
Walsh system in Lebesgue spaces, Math. Inequal. Appl., 27, no. 1, (2024), 137-147.
[3] D. Baramidze, N. Nadirashvili, L.-E. Persson and G. Tephnadze, Some weak-type in-
equalities and almost everywhere convergence of Vilenkin-Nörlund means, J. Inequal.
Appl., (2023), Paper No. 38, 14 pp.
[4] L. Baramidze, L. E. Persson, G. Tephnadze and P. Wall, Sharp Hp − Lp type inequal-
ities of weighted maximal operators of Vilenkin-Nörlund means and its applications,
J. Inequal. Appl., (2016), Paper No. 242, 20 pp.
 APPROXIMATION BY T MEANS 15

[5] I. Blahota and K. Nagy, Approximation by matrix transform of Vilenkin-Fourier

series, Publ. Math. Debrecen, 99, 1-2 (2021), 223-242.
[6] I. Blahota, K. Nagy and G. Tephnadze, Approximation by Marcinkiewicz Θ-means
of double Walsh-Fourier series, Math. Inequal. Appl., 22, 3 (2019) 837-853.
[7] I. Blahota and G. Tephnadze, On the (C, α)-means with respect to the Walsh system,
Anal. Math., 40 (2014), 161-174.
[8] I. Blahota, G. Tephnadze and R. Toledo, Strong convergence theorem of (C, α)-means
with respect to the Walsh system, Tohoku Math. J., 67, 4 (2015), 573-584.
[9] I. Blahota, L.E. Persson and G. Tephnadze, On the Nörlund means of Vilenkin-Fourier
series, Czech. Math J., 65, 4 (2015), 983-1002.
[10] I. Blahota and K. Nagy, Approximation by Θ-means of Walsh-Fourier series, Anal.
Math. 44, 1 (2018), 57-71.
[11] I. Blahota, K. Nagy and G. Tephnadze, Approximation by Marcinkiewicz Θ-means
of double Walsh-Fourier series, Math. Inequal. Appl., 22, 3 (2019), 837-853.
[12] S. Fridli, On the rate of convergence of Cesaro means of Walsh-Fourier series, J.
Approx. Theory, 76, 1 (1994), 31-53.
[13] S. Fridli, P, Manchanda and A. Siddiqi, Approximation by Walsh-Nörlund means,
Acta Sci. Math., 74 (3-4), 593-608.
[14] G. Gát, Cesàro means of integrable functions with respect to unbounded Vilenkin
systems. J. Approx. Theory, 124, 1 (2003), 25-43.
[15] U. Goginava, Maximal operators of Walsh-Nörlund means on the dyadic Hardy
spaces, Acta Math, Hungar., 169, 1 (2023), 171-190.
[16] U. Goginava, Marcinkiewicz-Fejer means of d-dimensional Walsh-Fourier series, 307,
1 (2015), 206 - 218.
[17] B. I. Golubov, A. V. Efimov and V. A. Skvortsov, Walsh series and transforms,
Kluwer Academic Publishers Group, Dordrecht, 1991.
[18] N. Memić, An estimate of the maximal operator of the Nörlund logarithmic means
with respect to the character system of the group of 2-adic integers on the Hardy
space H1 , Bull. Iranian Math. Soc., 48, 6 (2022), 3381-3391.
[19] F. Móricz and B. E. Rhoades, Approximation by weighted means of Walsh-Fourier
series, Internat. J. Math. Math. Sci., 19, no 1, (1996), 1-8.
[20] F. Móricz and A. Siddiqi, Approximation by Nörlund means of Walsh-Fourier series,
J. Approx. Theory, 70, 3 (1992), 375-389.
[21] K. Nagy, Approximation by Nörlund means of double Walsh-Fourier series for Lips-
chitz functions, Math. Inequal. Appl., 15, 2 (2012), 301-322.
[22] K. Nagy, Approximation by Nörlund means of Walsh-Kaczmarz-Fourier series. Geor-
gian Math. J. 18, 1 (2011), 147-162.
[23] K. Nagy, Approximation by Nörlund means of quadratical partial sums of double
Walsh-Fourier series, Anal. Math., 36, 4 (2010), 299-319.
[24] K. Nagy and G. Tephnadze, Walsh-Marcinkiewicz means and Hardy spaces, Cent.
Eur. J. Math., 12, 8 (2014), 1214-1228.
[25] K. Nagy and G. Tephnadze, Approximation by Walsh-Marcinkiewicz means on the
Hardy space , Kyoto J. Math., 54, 3 (2014), 641-652.
[26] K. Nagy and G. Tephnadze, Strong convergence theorem for Walsh-Marcinkiewicz
means, Math. Inequal. Appl., 19, 1 (2016), 185-195.
[27] K. Nagy and G. Tephnadze, The Walsh-Kaczmarz-Marcinkiewicz means and Hardy
spaces, Acta Math. Hungar., 149, 2 (2016), 346-374.
[28] J. Pál and P. Simon, On a generalization of the concept of derivative, Acta Math.
Hungar., 29 (1977), 155-164.
[29] L. E. Persson, G. Tephnadze and F. Weisz, Martingale Hardy Spaces and Summability
of Vilenkin-Fourier Series, Birkhäuser-Springer, Cham., 2022.
[30] F. Schipp, Certain rearranngements of series in the Walsh series, Mat. Zametki, 18
(1975), 193-201.
16 N. ANAKIDZE, N. ARESHIDZE, L.-E. PERSSON AND G. TEPHNADZE

[31] F. Schipp, W. R. Wade, P. Simon and J. Pál, Walsh Series. An Introduction to Dyadic
Harmonic Analysis, Adam Hilger, Ltd., Bristol, 1990.
[32] P Simon, Strong convergence of certain means with respect to the Walsh-Fourier
series, Acta Math. Hungar., 49, 3-4 (1987), 425-431.
[33] P Simon, On the Cesáro summability with respect to the Walsh-Kaczmarz system,
J. Approx. Theory, 106, 2 (2000), 249-261.
[34] G. Tutberidze, Maximal operators of means with respect to the Vilenkin system,
Nonlinear Stud., 27, no. 4, (2020), 1-11.
[35] N. Y. Vilenkin, On a class of complete orthonormal systems, Amer. Math. Soc.
Transi., 28, (2) (1963), 1-35.
[36] N. Y. Vilenkin, On the theory of lacunary orthogonal systems, (Russian) Izv. Akad.
Nauk SSSR, Ser. Mat., 13 (1949), 245-252.
[37] N. Y. Vilenkin, On the theory of Fourier integrals on topological groups, Mat. Sbornik
N. S., 30, (72) (1952), 233-244.
[38] F. Weisz, Martingale Hardy Spaces and their Applications in Fourier Analysis,
Springer, Berlin-Heideiberg-New York, 1994.
[39] F. Weisz, Hardy spaces and Cesàro means of two-dimensional Fourier series, Bolyai
Soc. Math. Studies, (1996), 353-367.
[40] A. Zygmund, Trigonometric Series, Vol. 1, Cambridge Univ. Press, 1959.

N. Anakidze, The University of Georgia, School of Science and Technol-

ogy, 77a Merab Kostava St, Tbilisi, 0171, Georgia.
Email address: [email protected]

N. Areshidze, Tbilisi State University, Faculty of Exact and Natural Sci-

ences, Department of Mathematics, Chavchavadze str. 1, tbilisi 0128, georgia
Email address: [email protected]

L.-E. Persson, UiT The Arctic University of Norway, P.O. Box 385, N-
8505, Narvik, Norway and department of Mathematics and Computer Science,
Karlstad University, 65188 Karlstad, Sweden.
Email address: [email protected]

G. Tephnadze, The University of Georgia, School of Science and Tech-

nology, 77a Merab Kostava St, Tbilisi, 0171, Georgia.
Email address: [email protected]