Alex Vilenkin (Tufts University) Miami, December 2014 TOPOLOGICAL DEFECTS FROM THE MULTIVERSE.
Some inequalities for Cesàro means of double Vilenkin...
Transcript of Some inequalities for Cesàro means of double Vilenkin...
Tepnadze and Persson Journal of Inequalities and Applications (2018) 2018:352 https://doi.org/10.1186/s13660-018-1929-y
R E S E A R C H Open Access
Some inequalities for Cesàro means ofdouble Vilenkin–Fourier seriesT. Tepnadze1 and L.E. Persson1*
*Correspondence:[email protected] Artic University of Norway,Narvik, Norway
AbstractIn this paper, we state and prove some new inequalities related to the rate of Lp
approximation by Cesàro means of the quadratic partial sums of doubleVilenkin–Fourier series of functions from Lp.
MSC: 42C10; 42B25
Keywords: Inequalities; Approximation; Vilenkin system; Vilenkin–Fourier series;Cesàro means; Convergence in norm
1 IntroductionLet N+ denote the set of positive integers, and let N := N+ ∪ {0}. Let m := (m0, m1, . . .) be asequence of positive integers not less than 2. Denote by Zmk := {0, 1, . . . , mk –1} the additivegroup of integers modulo mk . Define the group Gm as the complete direct product of thegroups Zmj with the product of the discrete topologies of Zmj .
The direct product of the measures
μk({j}) :=
1mk
(j ∈ Zmk )
is the Haar measure on Gm with μ(Gm) = 1. If the sequence m is bounded, then Gm iscalled a bounded Vilenkin group. In this paper, we consider only bounded Vilenkin groups.The elements of Gm can be represented by sequences x := (x0, x1, . . . , xj, . . .) (xj ∈ Zmj ). Thegroup operation + in Gm is given by
x + y =((x0 + y0) mod m0, . . . , (xk + yk) mod mk , . . .
)
for x := (x0, . . . , xk , . . .) and y := (y0, . . . , yk , . . .) ∈ Gm. The inverse of + will be denoted by –.It is easy to give a base for the neighborhoods of Gm:
I0(x) := Gm,
In(x) := {y ∈ Gm|y0 = x0, . . . , yn–1 = xn–1}
for x ∈ Gm and n ∈ N . Define In := In(0) for n ∈ N+. Set en := (0, . . . , 0, 1, 0, . . .) ∈ Gm, wherethe nth coordinate of which is 1, and the rest are zeros (n ∈ N ).
© The Author(s) 2018. This article is distributed under the terms of the Creative Commons Attribution 4.0 International License(http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in anymedium, pro-vided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, andindicate if changes were made.
Tepnadze and Persson Journal of Inequalities and Applications (2018) 2018:352 Page 2 of 17
We define the so-called generalized number system based on m as follows: M0 := 1,Mk+1 := mkMk (k ∈ N ). Then every n ∈ N can be uniquely expressed as n =
∑∞j=0 njMj,
where nj ∈ Zmj (j ∈ N+), and only a finite number of nj differ from zero. We also use thefollowing notation: |n| := max{k ∈ N : nk �= 0} (i.e., M|n| ≤ n < M|n|+1, n �= 0). For x ∈ Gm, wedenote |x| :=
∑∞j=0
xjMj+1
(xj ∈ Zmj ).Next, we introduce on Gm an orthonormal system, which is called the Vilenkin sys-
tem. First, we define the complex-valued functions rk(x) : Gm → C, the generalizedRademacher functions, as follows:
rk(x) := exp2π ixk
mk
(i2 = –1, x ∈ Gm, k ∈ N
).
Now we define the Vilenkin system ψ := (ψn : n ∈ N) on Gm as
ψn(x) :=∞∏
k=0
rnkk (x) (n ∈ N).
In particular, if m = 2, then we call this system the Walsh–Paley system. Each ψn is acharacter of Gm, and all characters of Gm are of this norm. Moreover, ψn(–x) = ψ̄n(x).
The Dirichlet kernels are defined by
Dn :=n–1∑
k=0
ψk (n ∈ N+).
Recall that (see [20] or [23])
DMn (x) =
⎧⎨
⎩Mn if x ∈ In,
0 if x /∈ In.(1)
The Vilenkin system is orthonormal and complete in L1(Gm) (see [1]).Next, we introduce some notation from the theory of two-dimensional Vilenkin system.
Let m̃ be a sequence like m. The relation between the sequences (m̃n) and (M̃n) is the sameas between sequences (mn) and (Mn). The group Gm × Gm̃ is called a two-dimensionalVilenkin group. The normalized Haar measure is denoted by μ as in the one-dimensionalcase. We also suppose that m = m̃ and Gm × Gm̃ = G2
m.The norm of the space Lp(G2
m) is defined by
‖f ‖p :=(∫
G2m
∣∣f (x, y)∣∣p dμ(x, y)
)1/p
(1 ≤ p < ∞).
Denote by C(G2m) the class of continuous functions on the group G2
m endowed with thesupremum norm.
For brevity in notation, we write L∞(G2m) instead of C(G2
m).The two-dimensional Fourier coefficients, the rectangular partial sums of the Fourier
series, and the Dirichlet kernels with respect to the two-dimensional Vilenkin system are
Tepnadze and Persson Journal of Inequalities and Applications (2018) 2018:352 Page 3 of 17
defined as follows:
f̂ (n1, n2) :=∫
G2m
f (x, y)ψ̄n1 (x)ψ̄n2 (y) dμ(x, y),
Sn1,n2 (x, y, f ) :=n1–1∑
k1=0
n2–1∑
k2=0
f̂ (k1, k2)ψk1 (x)ψk2 (y),
Dn1,n2 (x, y) := Dn1 (x)Dn2 (y),
Denote
S(1)n (x, y, f ) :=
n–1∑
l=0
f̂ (l, y)ψ̄l(x),
S(2)m (x, y, f ) :=
m–1∑
r=0
f̂ (x, r)ψ̄r(y),
where
f̂ (l, y) =∫
Gm
f (x, y)ψl(x) dμ(x)
and
f̂ (x, r) =∫
Gm
f (x, y)ψr(y) dμ(y).
The (C, –α) means of the double Vilenkin–Fourier series are defined as follows:
σ –αn (f , x, y) =
1A–α
n–1
n∑
j=1
A–α–1n–j Sj,j(f , x, y),
where
Aα0 = 1, Aα
n =(α + 1) · · · (α + n)
n!.
It is well known that (see [28])
Aαn =
n∑
k=0
Aα–1k , (2)
Aαn – Aα
n–1 = Aα–1n , (3)
and
c1(α)nα ≤ Aαn ≤ c2(α)nα , (4)
where positive constants c1 and c2 depend on α.
Tepnadze and Persson Journal of Inequalities and Applications (2018) 2018:352 Page 4 of 17
The dyadic partial moduli of continuity of a function f ∈ Lp(G2m) in the Lp-norm are
defined by
ω1
(f ,
1Mn
)
p= sup
u∈In
∥∥f (· + u, ·) – f (·, ·)∥∥p
and
ω2
(f ,
1Mn
)
p= sup
v∈In
∥∥f (·, · + v) – f (·, ·)∥∥p,
whereas the dyadic mixed modulus of continuity is defined as follows:
ω1,2
(f ,
1Mn
,1
Mm
)
p
= sup(u,v)∈In×Im
∥∥f (· + u, · + v) – f (· + u, ·) – f (·, · + v) + f (·, ·)∥∥p.
It is clear that
ω1,2
(f ,
1Mn
,1
Mm
)
p≤ ω1
(f ,
1Mn
)
p+ ω2
(f ,
1Mm
)
p.
The dyadic total modulus of continuity is defined by
ω
(f ,
1Mn
)
p= sup
(u,v)∈In×In
∥∥f (· + u, · + v) – f (·, ·)∥∥p.
The problems of summability of partial sums and Cesàro means for Walsh–Fourier serieswere studied in [2, 13–19, 21, 22, 25, 26].
The convergence issue of Fejér (and Cesàro) means on the Walsh and Vilenkin groupsfor unbounded case were studied in [3–11].
In his monograph [27], Zhizhinashvili investigated the behavior of Cesàro (C,α)-meansfor double trigonometric Fourier series in detail. Goginava [18] studied the analogousquestion in the case of the Walsh system. In particular, the following theorems wereproved.
Theorem A Let f belong to Lp(G2) for some p ∈ [1,∞] and α ∈ (0, 1). Then, for any 2k ≤n < 2k+1 (k, n ∈ N ), we have the inequality
∥∥σ –α
2k (f ) – f∥∥
p ≤ c(α)
{
2kαω1(f , 1/2k–1)
p + 2kαω2(f , 1/2k–1)
p
+k–2∑
r=0
2r–kω1(f , 1/2r)
p +k–2∑
s=0
2s–kω2(f , 1/2s)
p
}
.
Tepnadze and Persson Journal of Inequalities and Applications (2018) 2018:352 Page 5 of 17
Theorem B Let f belong to Lp(G2) for some p ∈ [1,∞] and α ∈ (0, 1). Then, for any 2k ≤n < 2k+1 (k, n ∈ N ), we have the inequality
∥∥σ –αn (f ) – f
∥∥p ≤ c(α)
{
2kαkω1(f , 1/2k–1)
p + 2kαkω2(f , 1/2k–1)
p
+k–2∑
r=0
2r–kω1(f , 1/2r)
p +k–2∑
s=0
2s–kω2(f , 1/2s)
p
}
.
In this paper, we state and prove analogous results in the case of double Vilenkin–Fourierseries. Our main results are the following theorems.
Theorem 1 Let f belong to Lp(G2m) for some p ∈ [1,∞] and α ∈ (0, 1). Then, for any Mk ≤
n < Mk+1 (k, n ∈ N ), we have the inequality
∥∥σ –α
Mk(f ) – f
∥∥
p ≤ c(α)
(
ω1(f , 1/Mk–1)pMαk + ω2(f , 1/Ml–1)pMα
k
+k–2∑
r=0
Mr
Mkω1(f , 1/Mr)p +
k–2∑
s=0
Ms
Mkω2(f , 1/Ms)p
)
.
Theorem 2 Let f belong to Lp(G2m) for some p ∈ [1,∞] and α ∈ (0, 1). Then, for any Mk ≤
n < Mk+1 (k, n ∈ N ), we have the inequality
∥∥σ –αn (f ) – f
∥∥p
≤ c(α)
(
ω1(f , 1/Mk–1)pMαk log n + ω2(f , 1/Ml–1)pMα
k log n
+k–2∑
r=0
Mr
Mkω1(f , 1/Mr)p +
k–2∑
s=0
Ms
Mkω2(f , 1/Ms)p
)
.
To make the proofs of these theorems clearer, we formulate some auxiliary lemmas inSect. 2. Some of these lemmas are new and of independent interest. Detailed proofs canbe found in Sect. 3.
2 Auxiliary lemmasTo prove Theorems 1 and 2, we need the following three lemmas (see [1, 12], and [8],respectively)
Lemma 1 Let α1,α2, . . . ,αn be real numbers. Then
1n
∫
G
∣∣∣∣∣
n∑
k=1
αkDk(x)
∣∣∣∣∣dμ(x) ≤ c√
n
( n∑
k=1
α2k
)1/2
.
Lemma 2 Let α1,α2, . . . ,αn be real numbers. Then
1n
∫
G2m
∣∣∣∣∣
n∑
k=1
αkDk(x)Dk(y)
∣∣∣∣∣dμ(x, y) ≤ c√
n
( n∑
k=1
α2k
)1/2
.
Tepnadze and Persson Journal of Inequalities and Applications (2018) 2018:352 Page 6 of 17
Lemma 3 Let 0 ≤ j < nsMs and 0 ≤ ns < ms. Then
DnsMs–j = DnsMs – ψnsMs–1D̄j.
We also need the following new nemmas of independent interest.
Lemma 4 Let f belong to Lp(G2m) for some p ∈ [1,∞]. Then, for every α ∈ (0, 1), we have
the inequality
I :=1
A–αn
∥∥∥∥∥
∫
G2m
Mk–1∑
i=1
A–α–1n–i Di(u)Di(v)
[f (· – u, ·–) – f (·, ·)]dμ(u, v)
∥∥∥∥∥
p
≤k–2∑
r=0
Mr
Mkω1(f , 1/Mr)p +
k–2∑
s=0
Ms
Mkω2(f , 1/Ms)p,
where Mk ≤ n < Mk+1.
Lemma 5 Let α ∈ (0, 1) and p = Mk , Mk + 1, . . . . Then
II :=∫
G2m
∣∣∣∣∣
Mk∑
i=1
A–α–1p–i Di(u)Di(v)
∣∣∣∣∣dμ(u, v) ≤ c(α) < ∞, k = 1, 2, . . . .
Lemma 6 We have the inequality
III :=∫
G2m
∣∣∣∣∣
n∑
i=1
A–α–1n–i Di(u)Di(v)
∣∣∣∣∣dμ(u, v) ≤ c(α) log n
3 The detailed proofs
Proof of Lemma 3 Applying Abel’s transformation, from (2) we get
I ≤ 1A–α
n
∥∥∥∥∥
∫
G2m
Mk–1–1∑
i=1
A–α–2n–i
i∑
l=1
Di(u)Di(v)[f (· – u, · – v) – f (·, ·)]dμ(u, v)
∥∥∥∥∥
p
+1
A–αn
∥∥∥∥∥
∫
G2m
A–α–1n–Mk–1
Mk–1∑
i=1
Di(u)Di(v)[f (· – u, · – v) – f (·, ·)]dμ(u, v)
∥∥∥∥∥
p
:= I1 + I2, (5)
where the first and second terms on the right side of inequality (5) are denoted by I1 andI2, respectively.
Tepnadze and Persson Journal of Inequalities and Applications (2018) 2018:352 Page 7 of 17
For I2, we have the estimate
I2 ≤ 1A–α
n
∥∥∥∥∥
∫
G2m
A–α–1n–Mk–1
k–2∑
r=1
Mr+1–1∑
i=Mr
Di(u)Di(v)
× [f (· – u, · – v) – f (·, ·)]
∥∥∥∥∥
p
dμ(u, v)
≤ 1A–α
n
∥∥∥∥∥
∫
G2m
A–α–1n–Mk–1
k–2∑
r=1
Mr+1–1∑
i=Mr
Di(u)Di(v)
× [f (· – u, · – v) – SMr ,Mr (· – u, · – v, f )
]dμ(u, v)
∥∥∥∥∥
p
+1
A–αn
∥∥∥∥∥
∫
G2m
A–α–1n–Mk–1
k–2∑
r=1
Mr+1–1∑
i=Mr
Di(u)Di(v)
× [SMr ,Mr (· – u, · – v, f ) – SMr ,Mr (·, ·, f )
]dμ(u, v)
∥∥∥∥∥
p
+1
A–αn
∥∥∥∥∥
∫
G2m
A–α–1n–Mk–1
k–2∑
r=1
Mr+1–1∑
i=Mr
Di(u)Di(v)
× [SMr ,Mr (·, ·, f ) – f (·, ·)]dμ(u, v)
∥∥∥∥∥
p
:= I21 + I22 + I23, (6)
where the first, second, and third terms on the right side of inequality (6) are denoted byI21, I22, and I23, respectively.
It is evident that
∫
G2m
Mr+1–1∑
i=Mr
Di(u)Di(v)[SMr ,Mr (· – u, · – v, f ) – SMr ,Mr (·, ·, f )
]dμ(u, v)
=Mr+1–1∑
i=Mr
(∫
G2m
Di(u)Di(v)SMr ,Mr (· – u, · – v, f ) dμ(u, v) – SMr ,Mr (·, ·, f ))
=Mr+1–1∑
i=Mr
(Si
(·, ·, SMr ,Mr (f ))
– SMr ,Mr (·, ·, f ))
=Mr+1–1∑
i=Mr
(SMr ,Mr (·, ·, f ) – SMr ,Mr (·, ·, f )
)= 0.
Hence
I22 = 0. (7)
Tepnadze and Persson Journal of Inequalities and Applications (2018) 2018:352 Page 8 of 17
Moreover, by the generalized Minkowski inequality, Lemma 2, and by (1) and (4) weobtain
I21 ≤ 1A–α
n
∣∣A–α–1n–Mk–1
∣∣k–2∑
r=1
∫
G2m
∣∣∣∣∣
Mr+1–1∑
i=Mr
Di(u)Di(v)
∣∣∣∣∣
× ∥∥f (· – u, · – v) – SMr ,Mr (· – u, · – v, f )∥∥
p dμ(u, v)
≤ c(α)Mk
k–2∑
r=1
(ω1(f , 1/Mr)p + ω2(f , 1/Mr)p
)
×∫
G2m
∣∣∣∣∣
Mr+1–1∑
i=Mr
Di(x)Di(y)
∣∣∣∣∣dμ(u, v)
≤ c(α)k–2∑
r=1
Mr
Mk
(ω1(f , 1/Mr)p + ω2(f , 1/Mr)p
). (8)
The estimation of I23 is analogous to that of I21:
I23 ≤ c(α)k–2∑
r=1
Mr
Mk
(ω1(f , 1/Mr)p + ω2(f , 1/Mr)p
). (9)
Analogously, we can estimate I1 as follows:
I1 ≤ 1A–α
n
k–2∑
r=1
∥∥∥∥∥
∫
G2m
Mr+1–1∑
i=Mr
A–α–2n–i
i∑
l=1
Dl(u)Dl(v)
× [f (· – u, · – v) – SMr ,Mr (· – u, · – v, f )
]dμ(u, v)
∥∥∥∥∥
p
+1
A–αn
k–2∑
r=1
∥∥∥∥∥
∫
G2m
Mr+1–1∑
i=Mr
A–α–2n–i
i∑
l=1
Dl(u)Dl(v)
× [SMr ,Mr (· – u, · – v, f ) – SMr ,Mr (·, ·, f )
]∥∥∥∥∥
p
dμ(u, v)
+1
A–αn
k–2∑
r=1
∥∥∥∥∥
∫
G2m
Mr+1–1∑
i=Mr
A–α–2n–i
i∑
l=1
Dl(u)Dl(v)
× [SMr ,Mr (·, ·, f ) – f (·, ·)]dμ(u, v)
∥∥∥∥∥
p
≤ 1A–α
n
k–2∑
r=1
∫
G2m
∣∣∣∣∣
Mr+1–1∑
i=Mr
A–α–2n–i
i∑
l=1
Dl(u)Dl(v)
∣∣∣∣∣
× ∥∥f (· – u, · – v) – SMr ,Mr (· – u, · – v, f )∥∥
p dμ(u, v)
Tepnadze and Persson Journal of Inequalities and Applications (2018) 2018:352 Page 9 of 17
+1
A–αn
k–2∑
r=1
∫
G2m
∣∣∣∣∣
Mr+1–1∑
i=Mr
A–α–2n–i
i∑
l=1
Dl(u)Dl(v)
∣∣∣∣∣
× ∥∥SMr ,Mr (·, ·, f ) – f (·, ·)∥∥p dμ(u, v)
≤ c(α)Mαk
k–2∑
r=1
Mr+1–1∑
i=Mr
(n – i)–α–2i(ω1(f , 1/Mr)p + ω2(f , 1/Mr)p
)
≤ c(α)Mαk
k–2∑
r=1
Mr+1–1∑
i=Mr
(n – Mr+1 – 1)–α–2i(ω1(f , 1/Mr)p + ω2(f , 1/Mr)p
)
≤ c(α)k–2∑
r=0
Mr
Mk
(ω1(f , 1/Mr)p + ω2(f , 1/Mr)p
). (10)
By combining (7)–(9) with (10) for I we find that
I ≤ c(α)k–2∑
r=0
Mr
Mk
(ω1(f , 1/Mr)p + ω2(f , 1/Mr)p
). (11)
The proof of Lemma 3 is complete. �
Proof of Lemma 4 It is evident that
II ≤∫
G2m
∣∣∣∣∣
Mk –1∑
i=1
A–α–1p–Mk +iDMk –i(u)DMk –i(v)
∣∣∣∣∣dμ(u, v)
+∣∣A–α–1
p–Mk
∣∣∫
G2m
DMk (u)DMk (v) dμ(u, v)
:= II1 + II2, (12)
where the first and second terms on the right side of inequality (12) are denoted by II1 andII2, respectively.
From (1) by |A–α–1p–Mk
| ≤ 1 we get that
II2 ≤ 1. (13)
Moreover, by Lemma 3 we have that
II1 ≤∫
G2m
∣∣∣∣∣
Mk –1∑
i=1
A–α–1p–Mk +iD̄i(u)D̄i(v)
∣∣∣∣∣dμ(u, v)
+∫
G2m
DMk (u)
∣∣∣∣∣
Mk –1∑
i=1
A–α–1p–Mk +iD̄i(v)
∣∣∣∣∣dμ(u, v)
+∫
G2m
DMk (v)
∣∣∣∣∣
Mk –1∑
i=1
A–α–1p–Mk +iD̄i(u)
∣∣∣∣∣dμ(u, v)
Tepnadze and Persson Journal of Inequalities and Applications (2018) 2018:352 Page 10 of 17
+
∣∣∣∣∣
Mk –1∑
i=1
A–α–1p–Mk +i
∣∣∣∣∣
∫
G2m
DMk (u)DMk (v) dμ(u, v)
:= II11 + II12 + II13 + II14, (14)
where the first, second, third, and fourth terms on the right side of inequality (14) aredenoted by II11, II12, II13, and II14 respectively.
From (1) and (4) it follows that
II14 ≤ c(α)∞∑
v=1
v–α–1 < ∞. (15)
By Applying Abel’s transformation, in view of Lemma 2, we have that
II11 ≤∫
G2m
∣∣∣∣∣
Mk –2∑
i=1
A–α–2p–Mk +i
i∑
l=1
D̄l(u)D̄l(v)
∣∣∣∣∣dμ(u, v)
+∫
G2m
∣∣∣∣∣A–α–1
p–1
Mk –1∑
i=1
D̄i(u)D̄i(v)
∣∣∣∣∣dμ(u, v)
≤ c(α)
{Mk –2∑
v=1
(p – Mk + i)–α–2i + (p – 1)–α–1Mk
}
≤ c(α)
{ ∞∑
i=1
i–α–1 + M–αk
}
< ∞. (16)
The estimation of II12 and II13 are analogous to the estimation of II11. Applying Abel’stransformation, in view of Lemma 1, we find that
II12 ≤∫
G2m
DMk (u)
∣∣∣∣∣
Mk –2∑
i=1
A–α–2p–Mk +i
i∑
l=1
D̄l(v)
∣∣∣∣∣dμ(u, v)
+∫
G2m
DMk (u)
∣∣∣∣∣A–α–1
p–1
Mk –1∑
i=1
D̄i(v)
∣∣∣∣∣dμ(u, v)
≤ c(α)
{Mk –2∑
v=1
(p – Mk + i)–α–2i + (p – 1)–α–1Mk
}
≤ c(α)
{ ∞∑
i=1
i–α–1 + M–αk
}
< ∞ (17)
and
III12 ≤∫
G2m
DMk (v)
∣∣∣∣∣
Mk –2∑
i=1
A–α–2p–Mk +i
i∑
l=1
D̄l(u)
∣∣∣∣∣dμ(u, v)
+∫
G2m
DMk (v)
∣∣∣∣∣A–α–1
p–1
Mk –1∑
i=1
D̄i(u)
∣∣∣∣∣dμ(u, v)
Tepnadze and Persson Journal of Inequalities and Applications (2018) 2018:352 Page 11 of 17
≤ c(α)
{Mk –2∑
v=1
(p – Mk + i)–α–2i + (p – 1)–α–1Mk
}
≤ c(α)
{ ∞∑
i=1
i–α–1 + M–αk
}
< ∞. (18)
The proof is complete by combining (12)–(18). �
Proof of Lemma 5 Let
n = nk1 Mk1 + · · · + nks Mks , k1 > · · · > ks ≥ 0.
Denote
n(i) = nki Mki + · · · + nks Mks , i = 1, 2, . . . , s.
Since (see [20])
Dj+nAMA = DnAMA + ψnAMA Dj, (19)
we find that
III ≤∫
G2m
∣∣∣∣∣
nk1 Mk1∑
i=1
A–α–1n–i Di(u)Di(v)
∣∣∣∣∣dμ(u, v)
+∫
G2m
∣∣∣∣∣
n(2)∑
i=1
A–α–1n(2)–iDi(u)Di(v)
∣∣∣∣∣dμ(u, v)
+∫
G2m
Dnk1 Mk1(u)Dnk1 Mk1
(v)
∣∣∣∣∣
n(2)∑
i=1
A–α–1n(2)–i
∣∣∣∣∣dμ(u, v)
+∫
G2m
Dnk1 Mk1(u)
∣∣∣∣∣
n(2)∑
i=1
A–α–1n(2)–iDi(v)
∣∣∣∣∣dμ(u, v)
+∫
G2m
Dnk1 Mk1(v)
∣∣∣∣∣
n(2)∑
i=1
A–α–1n(2)–iDi(u)
∣∣∣∣∣dμ(u, v)
:= III1 + III2 + III3 + III4 + III5, (20)
where the first, second, third, fourth, and fifth terms on the right side of inequality (20)are denoted by III1, III2, III3, III4, and III5, respectively.
By (1) we have that
III3 ≤ c(α). (21)
Moreover, since (see [24])∣∣∣∣∣
n∑
i=1
A–α–1n–i Di(u)
∣∣∣∣∣
= O(|u|α–1), (22)
Tepnadze and Persson Journal of Inequalities and Applications (2018) 2018:352 Page 12 of 17
for III4, we get that
III4 ≤∫
G2m
Dnk1 Mk1(u)|v|α–1 dμ(u, v)
≤∫
Gm
|v|α–1 dμ(v) =1α
< ∞. (23)
Analogously, we find that
III5 ≤∫
G2m
Dnk1 Mk1(v)|u|α–1 dμ(u, v)
≤∫
Gm
|u|α–1 dμ(v) =1α
< ∞. (24)
For r ∈ {0, . . . mA – 1} and 0 ≤ j < MA (see [20]), this yields that
Dj+rMA =
( r–1∑
q=0
ψqMA
)
DMA + ψ rMA
Dj.
Thus we have
∫
G2m
nk1 Mk1 –1∑
i=1
A–α–1n–i Di(u)Di(v) dμ(u, v)
≤∫
G2m
nk1 –1∑
r=0
Mk1 –1∑
i=0
A–α–1n–i–rMk1
Di+rMk1(u)Di+rMk1
(v) dμ(u, v)
≤∫
G2m
nk1 –1∑
r=0
Mk1 –1∑
i=0
A–α–1n–i–rMk1
( r–1∑
q=0
ψqMk1
)
DMk1(u)
×( r–1∑
q=0
ψqMk1
)
DMk1(v) dμ(u, v)
+∫
G2m
nk1 –1∑
r=0
Mk1 –1∑
i=0
A–α–1n–i–rMk1
( r–1∑
q=0
ψqMk1
)
DMk1(u)ψ r
MADi(v) dμ(u, v)
+∫
G2m
nk1 –1∑
r=0
Mk1 –1∑
i=0
A–α–1n–i–rMk1
ψ rMA
Di(u)
( r–1∑
q=0
ψqMk1
)
DMk1(v) dμ(u, v)
+∫
G2m
nk1 –1∑
r=0
Mk1 –1∑
i=0
A–α–1n–i–rMk1
ψ rMA
Di(u)ψ rMA
Di(v) dμ(u, v).
On the other hand, by (1) and (4) we obtain that
∫
G2m
A–α–1n–nk1 Mk1
Dnk1 Mk1(u)Dnk1 Mk1
(v) dμ(u, v) ≤ c(α).
Tepnadze and Persson Journal of Inequalities and Applications (2018) 2018:352 Page 13 of 17
Consequently, for III1, we have the estimate
III1 ≤∫
G2m
DMk1(u)DMk1
(v)
∣∣∣∣∣
nk1 –1∑
r=0
Mk1∑
i=1
A–α–1n–i–rMk1
∣∣∣∣∣dμ(u, v)
+∫
G2m
DMk1(u)
∣∣∣∣∣
nk1 –1∑
r=0
Mk1∑
i=1
A–α–1n–i–rMk1
Di(v)
∣∣∣∣∣dμ(u, v)
+∫
G2m
DMk1(v)
∣∣∣∣∣
nk1 –1∑
r=0
Mk1∑
i=1
A–α–1n–i–rMk1
Di(u)
∣∣∣∣∣dμ(u, v)
+∫
G2m
∣∣∣∣∣
nk1 –1∑
r=0
Mk1∑
i=1
A–α–1n–i–rMk1
Di(u)Di(v)
∣∣∣∣∣dμ(u, v) + c(α)
:= III11 + III12 + III13 + III14 + c(α), (25)
where the first, second, third, and fourth terms on the right side of inequality (25) aredenoted by III11, III12, III13, and III14, respectively.
From Lemma 4 we have that
III14 ≤ c(α). (26)
The estimation of III11 is analogous to that of III3, and we find that
III11 ≤ c(α). (27)
The estimation of III12 and III13 is analogous to that of III4, and we obtain that
III12 < ∞ (28)
and
III13 < ∞. (29)
After substituting (21) and (23)–(29) into (20), we conclude that
∫
G2m
∣∣∣∣∣
n∑
i=1
A–α–1n–i Di(u)Di(v)
∣∣∣∣∣dμ(u, v)
≤∫
G2m
∣∣∣∣∣
n(2)∑
i=1
A–α–1n(2)–iDi(u)Di(v)
∣∣∣∣∣dμ(u, v) + c(α)
≤ · · · ≤∫
G2m
∣∣∣∣∣
n(s)∑
i=1
A–α–1n(s)–iDi(u)Di(v)
∣∣∣∣∣dμ(u, v) + c(α)s
≤ c(α) + c(α)s ≤ c(α) log n.
The proof is complete. �
Tepnadze and Persson Journal of Inequalities and Applications (2018) 2018:352 Page 14 of 17
Now we are ready to prove the main results.
Proof of Theorem 1 It is evident that∥∥σ –α
Mk(f ) – f
∥∥p
≤ 1A–α
Mk –1
∥∥∥∥∥
∫
G2m
Mk–1∑
i=1
A–α–1Mk –iDi(u)Di(v)
[f (· – u, · – v) – f (·, ·)]dμ(u, v)
∥∥∥∥∥
p
+1
A–αMk –1
∥∥∥∥∥
∫
G2m
Mk∑
i=Mk–1+1
A–α–1Mk –iDi(u)Di(v)
[f (· – u, · – v) – f (·, ·)]dμ(u, v)
∥∥∥∥∥
p
:= I + II. (30)
From Lemma 5 it follows that
I ≤ c(α)k–2∑
r=0
Mr
Mk
(ω1(f , 1/Mr)p + ω2(f , 1/Mr)p
). (31)
Moreover, for II , we have the estimate
II ≤ 1A–α
Mk –1
∥∥∥∥∥
∫
G2m
Mk∑
i=Mk–1+1
A–α–1Mk –iDi(u)Di(v)
× [f (· – u, · – v) – S(1)
Mk–1(· – u, · – v, f )
]dμ(u, v)
∥∥∥∥∥
p
+1
A–αMk –1
∥∥∥∥∥
∫
G2m
Mk∑
i=Mk–1+1
A–α–1Mk –iDi(u)Di(v)
× [S(1)
Mk–1(· – u, · – v, f ) – f (·, ·)]dμ(u, v)
∥∥∥∥∥
p
:= II1 + II2, (32)
where the first and second terms on the right side of inequality (32) are denoted by II1 andII2, respectively.
In view of the generalized Minkowski inequality, by (4) and Lemma 5 we get that
II1 ≤ 1A–α
Mk –1
∫
G2m
∣∣∣∣∣
Mk∑
i=Mk–1+1
A–α–1Mk –iDi(u)Di(v)
∣∣∣∣∣
× ∥∥f (· – u, · – v) – S(1)
Mk–1(· – u, · – v, f )
∥∥
p dμ(u, v)
≤ c(α)Mαk ω1(f , 1/Mk–1)p. (33)
The estimation of II2 is analogous to that of II1, and we find that
II2 ≤ c(α)Mαk ω2(f , 1/Mk–1)p. (34)
Combining (30)–(34), we obtain the proof of Theorem 1. �
Tepnadze and Persson Journal of Inequalities and Applications (2018) 2018:352 Page 15 of 17
Proof of Theorem 2 It is evident that
∥∥σ –α
n (f ) – f∥∥
p
≤ 1A–α
n–1
∥∥∥∥∥
∫
G2m
Mk–1∑
i=1
A–α–1n–i Di(u)Di(v)
[f (· – u, · – v) – f (·, ·)]dμ(u, v)
∥∥∥∥∥
p
+1
A–αn–1
∥∥∥∥∥
∫
G2m
Mk∑
i=Mk–1+1
A–α–1n–i Di(u)Di(v)
[f (· – u, · – v) – f (·, ·)]dμ(u, v)
∥∥∥∥∥
p
+1
A–αn–1
∥∥∥∥∥
∫
G2m
n∑
i=Mk +1
A–α–1n–i Di(u)Di(v)
[f (· – u, · – v) – f (·, ·)]dμ(u, v)
∥∥∥∥∥
p
:= I + II + III, (35)
where the first, second, and third terms on the right side of inequality (35) are denoted byI , II , and III , respectively.
From Lemma 4 it follows that
I ≤ c(α)k–2∑
r=0
Mr
Mk
(ω1(f , 1/Mr)p + ω2(f , 1/Mr)p
). (36)
Next, we repeat the arguments just in the same way as in the proof of Theorem 1 andfind that
II ≤ c(α)Mαk(ω1(f , 1/Mk–1)p + ω2(f , 1/Mk–1)p
). (37)
On the other hand, for III, we have
III ≤ 1A–α
n–1
∥∥∥∥∥
∫
G2m
n∑
i=Mk +1
A–α–1n–i Di(u)Di(v)
× [f (· – u, · – v) – f (·, ·)]
∥∥∥∥∥
p
dμ(u, v)
≤ 1A–α
n
∥∥∥∥∥
∫
G2m
n∑
i=Mk +1
A–α–1n–i Di(u)Di(v)
× [f (· – u, · – v) – SMk ,Mk (· – u, · – v, f )
]dμ(u, v)
∥∥∥∥∥
p
≤ 1A–α
n
∥∥∥∥∥
∫
G2m
n∑
i=Mk +1
A–α–1n–i Di(u)Di(v)
× [SMk ,Mk (· – u, · – v, f ) – SMk ,Mk (·, ·, f )
]dμ(u, v)
∥∥∥∥∥
p
Tepnadze and Persson Journal of Inequalities and Applications (2018) 2018:352 Page 16 of 17
≤ 1A–α
n
∥∥∥∥∥
∫
G2m
n∑
i=Mk +1
A–α–1n–i Di(u)Di(v)
× [SMk ,Mk (·, ·, f ) – f (·, ·)]dμ(u, v)
∥∥∥∥∥
p
:= III1 + III2 + III3, (38)
where the first, second, and third terms on the right side of inequality (38) are denoted byIII1, III2, and III3, respectively.
It is easy to show that
III2 = 0. (39)
By the generalized Minkowski inequality and Lemma 5, for III1, we obtain that
III1 ≤ 1A–α
n
∫
G2m
∣∣∣∣∣
n∑
i=Mk +1
A–α–1n–i Di(u)Di(v)
∣∣∣∣∣
× ∥∥f (· – u, · – v) – SMr ,Mr (· – u, · – v, f )∥∥
p dμ(u, v)
≤ c(α)Mαk(ω1(f , 1/Mk–1)p + ω2(f , 1/Mk–1)p
)
×∫
G2m
∣∣∣∣∣
n∑
v=Mk +1
A–α–1n–v Dv(u)Dv(v)
∣∣∣∣∣dμ(u, v)
≤ c(α)Mαk log n
(ω1(f , 1/Mk–1)p + ω2(f , 1/Mk–1)p
). (40)
The estimation of III3 is analogous to that of III2, and we find that
III3 ≤ c(α)Mαk log n
(ω1(f , 1/Mk–1)p + ω2(f , 1/Mk–1)p
). (41)
After substituting (36)–(37) and (41) into (35), we obtain the proof of Theorem 2. �
AcknowledgementsThe authors would like to thank the referees for helpful suggestions.
FundingNot applicable.
Competing interestsThe authors declare that they have no competing interests.
Authors’ contributionsThe authors contributed equally to the writing of this paper. Both authors approved the final version of the manuscript.
Publisher’s NoteSpringer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Received: 22 August 2018 Accepted: 2 December 2018
References1. Agaev, G.N., Vilenkin, N.Ya., Dzhafarli, G.M., Rubinshtejn, A.I.: Multiplicative Systems of Functions and Harmonic
Analysis on Zero-Dimensional Groups, Ehlm, Baku (1981) (in Russian)2. Fine, N.J.: Cesàro summability of Walsh–Fourier series. Proc. Natl. Acad. Sci. USA 41, 558–591 (1995)
Tepnadze and Persson Journal of Inequalities and Applications (2018) 2018:352 Page 17 of 17
3. Gát, G.: On the pointwise convergence of Cesàro means of two-variable functions with respect to unboundedVilenkin systems. J. Approx. Theory 128(1), 69–99 (2004)
4. Gát, G.: Almost everywhere convergence of Fejér means of L1 functions on rarely unbounded Vilenkin groups. ActaMath. Sin. Engl. Ser. 23(12), 2269–2294 (2007)
5. Gát, G.: On almost everywhere convergence of Fourier series on unbounded Vilenkin groups. Publ. Math. (Debr.)75(1–2), 85–94 (2009)
6. Gát, G.: Some convergence and divergence results with respect to summation of Fourier series on one- andtwo-dimensional unbounded Vilenkin groups. Ann. Univ. Sci. Bp. Rolando Eötvös Nomin., Sect. Comput. 33, 157–173(2010)
7. Gát, G., Blahota, I.: Norm summability of Nörlund logarithmic means on unbounded Vilenkin groups. Anal. TheoryAppl. 24(1), 1–17 (2008)
8. Gát, G., Goginava, U.: Almost everywhere convergence of (C,α)-means of quadratical partial sums of doubleVilenkin–Fourier series. Georgian Math. J. 13(3), 447–462 (2006)
9. Gát, G., Goginava, U.: A weak type inequality for the maximal operator of (C,α)-means of Fourier series with respectto the Walsh–Kaczmarz system. Acta Math. Hung. 125(1–2), 65–83 (2009)
10. Gát, G., Goginava, U.: Norm convergence of double Fourier series on unbounded Vilenkin groups. Acta Math. Hung.152(1), 201–216 (2017)
11. Gát, G., Goginava, U.: Norm convergence of logarithmic means on unbounded Vilenkin groups. Banach J. Math. Anal.12(2), 422–438 (2018)
12. Glukhov, V.A.: On the summability of multiple Fourier series with respect to multiplicative systems. Mat. Zametki 39,665–673 (1986) In Russian
13. Goginava, U.: On the uniform convergence of Walsh–Fourier series. Acta Math. Hung. 93(1–2), 59–70 (2001)14. Goginava, U.: On the approximation properties of Cesàro means of negative order of Walsh–Fourier series. J. Approx.
Theory 115(1), 9–20 (2002)15. Goginava, U.: Uniform convergence of Cesàro means of negative order of double Walsh–Fourier series. J. Approx.
Theory 124(1), 96–108 (2003)16. Goginava, U.: On Cesàro means of double trigonometric Fourier series. Mat. Zametki 74(4), 502–507 (2003) (Russian),
translation in Math. Notes 74(3–4), 477–482 (2003)17. Goginava, U.: Cesàro means of double Walsh–Fourier series. Anal. Math. 30(4), 289–304 (2004)18. Goginava, U.: Approximation properties of (C,α) means of double Walsh–Fourier series. Anal. Theory Appl. 20(1),
77–98 (2004)19. Goginava, U., Nagy, K.: On the maximal operator of Walsh–Kaczmarz–Fejér means. Czechoslov. Math. J. 61(3),
673–686 (2011)20. Golubov, B.I., Efimov, A.V., Skvortsov, V.A.: Walsh Series and Transforms. Nauka, Moscow (1987) [in Russian]; English
translation, Kluwer Academic, Dordrecht (1991)21. Nagy, K.: Approximation by Cesàro means of negative order of Walsh–Kaczmarz–Fourier series. East J. Approx. 16(3),
297–311 (2010)22. Schipp, F.: Über gewisse Maximaloperatoren. Ann. Univ. Sci. Bp. Rolando Eötvös Nomin., Sect. Math. 18, 189–195
(1975)23. Schipp, F., Wade, W.R., Simon, P.: Walsh Series, an Introduction to Dyadic Harmonic Analysis. Hilger, Bristol (1990)24. Shavardenidze, G.: On the convergence of Cesáro means of negative order of Vilenkin–Fourier series.
arXiv:1811.0836725. Simon, P., Weisz, F.: Weak inequalities for Cesàro and Riesz summability of Walsh–Fourier series. J. Approx. Theory
151(1), 1–19 (2008)26. Tevzadze, V.I.: Uniform (C,α)(–1≤ α ≤ 0) summability of Fourier series with respect to the Walsh–Paley system. Acta
Math. Acad. Paedagog. Nyházi. 22(1), 41–61 (2006)27. Zhizhiashvili, L.V.: Trigonometric Fourier Series and Their Conjugates, Revised and updated translation of “Some
problems of the theory of trigonometric Fourier series and their conjugate series” (in Russian) [Tbilis. Gos. Univ.,Tbilisi, 1993]; English translation: Kluwer Academic, Dordrecht (1996)
28. Zygmund, A.: Trigonometric Series, Vol. 1, Cambridge University Press, Cambridge (1959)