Editorial Board

Supported by NSFC


Honorary Editor General ZHOU GuangZhao (Zhou Guang Zhao)
Editor General ZHU ZuoYan Institute of Hydrobiology, CAS, China
Editor-in-Chief YUAN YaXiang Academy of Mathematics and Systems Science, CAS, China
Associate Editors-in-Chief
CHEN YongChuan Tianjin University, China GE LiMing Academy of Mathematics and Systems Science, CAS, China
SHAO QiMan The Chinese University of Hong Kong, China XI NanHua Academy of Mathematics and Systems Science, CAS, China
ZHANG WeiPing Nankai University, China

Members

BAI ZhaoJun
University of California, Davis, USA
CAO DaoMin
Academy of Mathematics and Systems
Science, CAS, China
CHEN XiaoJun
The Hong Kong Polytechnic University,
China
CHEN ZhenQing
University of Washington, USA
CHEN ZhiMing
Academy of Mathematics and Systems
Science, CAS, China
CHENG ChongQing
Nanjing University, China
DAI YuHong
Academy of Mathematics and Systems
Science, CAS, China
DONG ChongYing
University of California, Santa Cruz,
USA
DUAN HaiBao
Academy of Mathematics and Systems
Science, CAS, China
E WeiNan
Princeton University, USA
Peking University, China
FAN JianQing
Princeton University, USA
FENG Qi
Academy of Mathematics and Systems
Science, CAS, China
FU JiXiang
Fudan University, China
GAO XiaoShan
Academy of Mathematics and Systems
Science, CAS, China
GE GenNian
Zhejiang University, China
GUO XianPing
Sun Yat-sen University, China
HE XuMing
University of Michigan, USA
HONG JiaXing
Fudan University, China
HSU Elton P.
Northwestern University, USA
JI LiZhen
University of Michigan, USA
JING Bing-Yi
The Hong Kong University of Science
and Technology, China
LI JiaYu
University of Science and Technology
of China, China
LIN FangHua
New York University, USA
LIU JianYa
Shandong University, China
LIU KeFeng
University of California, Los Angeles, USA
Zhejiang University, China
LIU XiaoBo
Peking University, China
University of Notre Dame, USA
MA XiaoNan
University of Denis Diderot-Paris 7,
France
MA ZhiMing
Academy of Mathematics and Systems
Science, CAS, China
MOK NgaiMing
The University of Hong Kong, China
PUIG Lluis
CNRS, Institute of Mathematics of Jussieu,
France
QIN HouRong
Nanjing University, China
RINGEL Claus M.
University of Bielefeld, Germany
SHANG ZaiJiu
Academy of Mathematics and Systems
Science, CAS, China
SHEN ZhongMin
Indiana University-Purdue University
Indianapolis, USA
SHU Chi-Wang
Brown University, USA
SIU Yum-Tong
Harvard University, USA
SUN LiuQuan
Academy of Mathematics and Systems
Science, CAS, China
SUN XiaoTao
Academy of Mathematics and Systems
Science, CAS, China
TAN Lei
University of Angers, France
TANG ZiZhou
Beijing Normal University, China
TEBOULLE Marc
Tel Aviv University, Israel
WANG FengYu
Beijing Normal University, China
WANG HanSheng
Peking University, China

WANG YueFei
Academy of Mathematics and Systems
Science, CAS, China
WU SiJue
University of Michigan, USA
WU SiYe
The University of Hong Kong, China
XIAO Jie
Tsinghua University, China
XIN ZhouPing
The Chinese University of Hong Kong,
China
XU Fei
Capital Normal University, China

XU Feng
University of California, Riverside, USA
XU JinChao
Pennsylvania State University, USA
XU XiaoPing
Academy of Mathematics and Systems
Science, CAS, China
YAN Catherine H. F.
Texas A&M University, USA
YANG DaChun
Beijing Normal University, China
YE XiangDong
Sun Yat-sen University, China
YU XingXing
Georgia Institute of Technology, USA
ZHANG James J.
University of Washington, USA
ZHANG JiPing
Peking University, China
ZHANG Ping
Academy of Mathematics and Systems
Science, CAS, China
ZHANG PingWen
Peking University, China
ZHANG ShouWu
Columbia University, USA
ZHANG Xu
Sichuan University, China
ZHOU XiangYu
Academy of Mathematics and Systems
Science, CAS, China
ZHU XiPing
Sun Yat-sen University, China
ZONG ChuanMing
Peking University, China


f f a t S l a i r o t i d E
CHAI Zhao
chaizhao@scichina.org
YANG ZhiHua
zhihua@scichina.org
ZHANG RuiYan
zhangry@scichina.org
SCIENCE CHINA
Mathematics
.
ARTICLES
.
April 2013 Vol. 56 No. 4: 737744
doi: 10.1007/s11425-013-4579-z
c Science China Press and Springer-Verlag Berlin Heidelberg 2013 math.scichina.com www.springerlink.com
Tensor products of ideal codes over Hopf algebras
GARC

IA-RUBIRA J. M. & L

OPEZ-RAMOS J. A.

Department of Algebra and Analysis, University of Almeria, Almeria 04120, Spain

Email: jrubira@gmail.com, jlopez@ual.es
Received July 7, 2011; accepted June 5, 2012; published online March 1, 2013
Abstract We study indecomposable codes over a family of Hopf algebras introduced by Radford. We use
properties of Hopf algebras to show that tensors of ideal codes are ideal codes, extending the corresponding
result that was previously given in the case of Taft Hopf algebras and showing the dierences with that case.
Keywords Radford Hopf algebra, ideal code, tensor product of ideals
MSC(2010) 97E10, 16T05
Citation: Garca-Rubira J M, Lopez-Ramos J A. Tensor products of ideal codes over Hopf algebras. Sci China
Math, 2013, 56: 737744, doi: 10.1007/s11425-013-4579-z
1 Introduction
The more mathematical structure one can add to a system the more information and a better description
can be obtained. A clear example of this occurs with algebraic codes. Linearity on a code implies, among
other things that one has not to compute all distances between every pair of codewords. If the code is also
cyclic, then we get fast encoding/decoding, etc. As Berman observed in [1], cyclic codes and Reed-Muller
codes can be seen as ideals in the group ring KG (where K is a nite eld and G is a nite cyclic group).
This fact has led many authors to the study of codes from a point of view of Ring Theory (cf. [3] for
instance). Recently, Wood in [9] and [10] has stated the suitability of Frobenius rings when studing codes
over a nite ring. More precisely, he showed that a nite Frobenius ring is characterized by the fact of
allowing the Extension Theorem for linear codes. One example of such rings are nite-dimensional Hopf
algebras. In [4] the authors characterized all indecomposable codes over an important family of Hopf
algebras: Taft Hopf algebras. Codes in such a family of Hopf algebras are shown to be a concatenation
of cyclic codes in KZ
n
. An application of identifying the concatenation of (ideal) codes with an (ideal)
code in a bigger algebra is that we can describe easily duals and tensor products of such codes. To do
so, dual and tensor product of (ideal) codes are equipped with an additional module structure given by
the antipode and the comultiplication of the underlying Hopf algebra respectively. In [5] the authors
considered a larger family of Hopf algebras, known as Radford Hopf algebras, that contains the rst one.
They give a decomposition theorem relating these two families of Hopf algebras and using it, duals of
indecomposable ideal codes are characterized. These new results extend those in [4], showing that these
codes are also concatenation of cyclic codes in a bigger algebra.
Our aim in this paper is to study the behavior of tensor products of indecomposable ideal codes
in Radford Hopf algebras. We recall that a linear code is an error-correcting code where any linear

Corresponding author
738 Garca-Rubira J M et al. Sci China Math April 2013 Vol. 56 No. 4
combination of two codewords is again a codeword, i.e., a linear code is a vector subspace of the considered
vector space. The length of a code is given by the dimension of the vector space and the dimension of the
code is its dimension as a vector space. If in addition we have an underlying ring structure, as is the case,
then some codes can also be considered as ideals when satisfying ideal structure. As noted above, [1] is the
rst work where this new point of view appears and our aim is to give a detailed description of ideal codes
obtained from some ideal codes in Radford Hopf algebra. To this end we will give a generator matrix,
i.e., a matrix whose rows determine a basis for the code, and thus a basis to generate all codewords.
Error correcting capability can be derived from a parameter of the code called the minimum distance.
To dene it we need the notion of Hamming distance. The Hamming distance between two codewords is
the number of positions where the corresponding symbols are dierent. Then the minimum distance of a
code is dened as the minimum of the Hamming distance between two any codewords. If the minimum
distance of a linear code is d, then the code is able to detect up to d 1 errors of changed symbols into
a transmitted codeword and to correct
d1
2
of these errors.
The paper is structured as follows. The second section recalls the denitions of the Taft and Radford
Hopf algebras. We give a sketch of the proof of the theorem appearing in [5, Theorem 2.2] that relates
both families in order to understand the structure of the indecomposables in the family of Radford Hopf
algebras. Then we start the study of tensor products of indecomposable codes. We recall that in [4] we
got that the tensor product of two any indecomposables in the Taft Hopf algebra can be identied with
a direct sum of indecomposable ideal codes. In the case of tensor products of simple codes, we got a
semisimple ideal code. Now we detach the fact that semisimplicity of tensor product of simple ideal codes
is lost when considering the family of Radford Hopf algebras (3.5). For prerequisites on Hopf algebras
we refer the reader to [2] or [6].
2 Two families of Hopf algebras
Let n > 1 be an integer and let K be a eld containing a primitive n-th root of unity (char(K) n).
Then we can provide a structure of bialgebra to the free algebra 1 = Kg, x over the non-commutative
indeterminates g and x. The comultiplication is dened by (g) = g g, (x) = x g +1 x and the
counit is given by (g) = 1, (x) = 0. If we consider the ideal I of 1 I = g, x : g
n
= 1, xg = gx, x
n
= 0,
then quotient algebra, that we will denote by T
n
= 1/I, is a Hopf algebra (known as Taft Hopf algebra)
whose antipode is dened by S(g) = g
1
and S(x) = xg
1
. It can be shown that g
i
x
j
: 0 i, j < n
is a basis for T
n
and so dim(T
n
) = n
2
. As noted before, indecomposable ideal codes over T
n
were studied
in [4].
Now let p > 1 be a prime number and assume char(K) ,= p. If we consider the ideal of 1, I

=
g, x : g
pn
= 1, xg = gx, x
n
= (g
n
1), then the quotient algebra A
p
= 1/I

is also a Hopf algebra

with antipode given by S(g) = g
1
, S(x) = xg
1
(as above, now g and x denote the corresponding
projections of the original elements g and x in 1). We can also show that the set B = g
i
x
j
: 0 i
< pn, 0 j < n is a basis for the algebra A
p
and thus dim(A
p
) = pn
2
.
This family of Hopf algebras was introduced by Radford in [7] as an example of Hopf algebra H whose
Jacobson radical J(H) is not a Hopf ideal, i.e., the semisimple quotient algebra H/J(H) does not admit
a Hopf algebra structure making the canonical projection into a Hopf algebra map. We will refer to A
p
as Radford Hopf algebra.
The following result appearing in [5] relates the two families of Hopf algebras dened above. We include
a sketch of the proof here in order to understand the incoming results better.
Theorem 2.1 (See [5, Theorem 2.2]). Assume that K contains a primitive p-th root of unity and an
n-th root of . Then
A
p

= T
n
M
n
(K)
(p1)
M
n
(K)
as algebras.
Garca-Rubira J M et al. Sci China Math April 2013 Vol. 56 No. 4 739
Proof. The element e =
1
p

p1
k=0
g
kn
is a central idempotent in A
p
and it can be shown that A
p
e

= T
n
as algebras via the isomorphism g ge, x xe.
On the other hand, A
p
(1 e)

= M
n
(K)
(p1)
M
n
(K). To prove this one shows that the set f
i

p1
i=1
is a complete set of orthogonal central idempotents in A
p
(1 e), where
f
i
=
1
p

p1

j=0

ij
g
nj

(1 e), i = 1, . . . , p 1.
Then
A
p
(1 e) =
p1

i=1
(A
p
(1 e))f
i
=
p1

i=1
A
p
f
i
,
and A
p
f
i

= M
n
(K) as algebras for every i = 1, . . . , p 1 via the isomorphism
gf
i

pi
diag(1,
1
, . . . ,
(n1)
),
xf
i

0 1 0 0
0 0 1 0
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
0 0 0 1

pi
1 0 0 0

.
3 Tensor product of indecomposables
Our aim is to study the tensor product of indecomposables in A
p
. Thus, by the precedent theorem we
have to consider, on one hand, those in T
n
, and on the other, those corresponding to the matrix blocks.
Indecomposable codes in T
n
are discussed in [4] and we will refer to them as indecomposable Taft codes.
Through the next subsection we recall their structure.
3.1 Indecomposable Taft codes
The Jacobson radical of T
n
is J(T
n
) = x. Then the application f : T
n
= T
n
/J(T
n
) KZ
n
given by
f(g) g and f(x) = 0 denes an isomorphism of algebras. Therefore the set of primitive idempotents
of KZ
n
,

e
l
=
1
n

n1

i=0

li
g
i

: 0 l < n

denes the set of isomorphism classes of simple T

n
-modules X
l
: 0 l < n. If we denote u
l
= e
l
and
X
l
= T
n
e
l
= Ku
l
, then we get that gu
l
=
l
u
l
and xu
l
= 0. Thus the set of class of isomorphism classes
of projective indecomposable modules of T
n
is obtained by lifting this set of orthogonal idempotents to
one in T
n
and these will be given by the projective cover P
l
of X
l
. Since e
l
considered as an element in
T
n
provides a lifted idempotent of e
l
, then the above set is indeed a complete set of primitive orthogonal
idempotents in T
n
. Therefore the projective cover of the simple T
n
e
l
will be dened by P
l
= He
l
.
The set v
i
= x
i
e
l
: i = 0, . . . , n 1 is a basis of P
l
and the action of T
n
on the module P
l
is given by
g v
i
=
(l+i)
v
i
,
x v
i
= v
i+1
, i = 0, . . . , n 2,
x v
n1
= 0.
(3.1)
The only submodules of P
l
are N
l,j
= J(T
n
)
j
P
l
for j = 0, . . . , n since T
n
is a serial algebra and a
basis for such module N
l,j
is given by the set v
j
, . . . , v
n1
, which can be derived from (3.1) and so
dim(N
l,j
) = n j.
740 Garca-Rubira J M et al. Sci China Math April 2013 Vol. 56 No. 4
Then we get a composition series of P
l
,
0 = N
l,n
N
l,n1
N
l,2
N
l,1
N
l,0
= P
l
whose decomposition factors are N
l,k
/N
l,k+1

= S
l+k
for k = 0, . . . , n 1, where the quotient mod-
ule P
l
/N
l,j
is isomorphic to N
l+j,nj
. To show this, let us denote by [v] the class of v P
l
. Then
[v
0
], . . . , [v
j1
] is a basis of P
l
/N
l,j
. The module structure is given by the following actions:
g [v
i
] = [g v
i
] = [
(i+l)
v
i
] =
(i+l)
[v
i
], i = 0, . . . , j 1,
x [v
i
] = [x v
i
] = [v
i+1
], i = 0, . . . , j 2,
x [v
j1
] = [x v
j1
] = [v
j
] = [0].
Hence the map N
l+j,nj
P
l
/N
l,j
, v
ni
[v
ji
] for i = 1, . . . , j is an isomorphism of modules.
A characterization of the tensor product of two any indecomposable Taft codes is shown in the following
result.
Theorem 3.1 (See [4, Theorem 2]). N
i,nr
N
j,ns

=

m
l=1
N
i+jl,nt
, where m = minr, s and
t = maxr, s.
3.2 Product of two simples in the matrix part
We recall that every matrix block A
p
f
j

= M
n
(K) is generated by a simple S
j
, i.e., M
n
(K)

= S
(n)
j
for
j = 1, . . . , p 1. From the proof of Theorem 2.1 it follows that the action of the generators of A
p
over
the basis of these simples is as follows: given S
j
= Kv
0
, . . . , v
n1
(j = 1, . . . , p 1),
g v
i
=
pj

i
v
i
,
x v
i
= v
i+1
, i = 0, . . . , n 2,
x v
n1
= (
pj
1)v
0
.
Lemma 3.2. S
j
S
j

= S
(n)
k
as A
p
-modules with k
p
2j as A
p
-modules.
Proof. Let us write S
j
S
j
= Kz
r,s
: 0 r, s n 1 for z
r,s
= v
r
v
s
. Let us show the action of g
on z
r,s
:
g z
r,s
= g v
r
g v
s
=
r

pj
v
r

pj
v
s
=
(r+s)

2(pj)
z
r,s
.
So let us consider z
k,l
such that k +l
n
0 and let t
i
= x
i
z
k,l
for i = 0, . . . , n1, and let us calculate
the action of x and g on the elements of B = t
0
, . . . , t
n1
:
g t
i
= g (x
i
z
k,l
) =
i
x
i
(gz
k,l
)
=
i
x
i
(
(k+l)

2(pj)
z
k,l
)
=
i

2(pj)
x
i
z
k,l
=
i

2(pj)
t
i
, i = 0, . . . , n 1,
x t
i
= x (x
i
z
k,l
) = x
i+1
z
k,l
= t
i+1
, i = 0, . . . , n 2,
x t
n1
= x (x
n1
z
k,l
) = x
n
z
k,l
= (g
n
1) z
k,l
= (
2(pj)
1)z
k,l
= (
2(pj)
1)t
0
since
n
= . Thus if z
k,l
is such that k +l
n
0 we get, on one side, by the action of g on the elements of
B, that B is a basis since every t
i
, i = 0, . . . , n1, is an eigenvector associated with a dierent eigenvalue
and, on the other hand that Kt
0
, t
1
, . . . , t
n1

= S
2j
as A
p
-modules.
Let us show now that x
i
z
k,nk
: 0 i, k n1 is a basis for S
j
S
j
. So let us denote t
i
= x
i
z
k,nu
,
i = 0, . . . , n 1 and s
m
= x
m
z
v,nv
, m = 0, . . . , n 1. By the precedent Kt
0
, t
1
, . . . , t
n1

= S
2j
and
Ks
0
, s
1
, . . . , s
n1

= S
2j
. Then we will check that Kt
0
, t
1
, . . . , t
n1
Ks
0
, s
1
, . . . , s
n1
= 0. Otherwise
Kt
0
, t
1
, . . . , t
n1
= Ks
0
, s
1
, . . . , s
n1
since both are simple. Now Ks
i
, t
i
Ks
l
, t
l
= 0 since they
are subspaces attached to dierent eigenvalues by the action of g. Assume now that s
i
= t
i
. Then
Garca-Rubira J M et al. Sci China Math April 2013 Vol. 56 No. 4 741
x
ni
s
i
= x
ni
t
i
and thus x
n
z
u,nu
= x
n
z
v,nv
, which is equivalent to (
2(pj)
1)z
u,nu
=
(
2(pj)
1)z
v,nv
which yields a contradiction. Then, by a reasoning on the dimension of the vector
spaces we get that S
j
S
j

= S
(n)
k
, k = 2j(mod p) as A
p
-modules.
Lemma 3.3. S
j
S
k

= S
(n)
q
as A
p
-modules for j + k = p + q.
Proof. Let q be such that j + k = p + q and let z
r,s
= v
r
w
s
. Then
g z
r,s
= g v
r
g w
s
=
r

pj
v
r

pk
w
s
=
(r+s)

pj+pk
v
r
v
s
=
(r+s)

pj+pk
z
r,s
=
(r+s)

pjk

p
z
r,s
=
(r+s)

pq
z
r,s
.
Now let t
i
= x
i
z
r,s
, r + s
n
0, i = 0, . . . , n 1.
Let us show the actions of g and x on these elements,
g t
i
= g (x
i
z
r,s
) =
i
x
i
(g z
r,s
) =
i

pq
x
i
z
r,s
=
i

pq
t
i
,
x t
i
= x (x
i
z
r,s
) = x
i+1
z
r,s
= t
i+1
, i = 0, . . . , n 2,
x t
n1
= x (x
n1
z
r,s
) = x
n
z
r,s
= (g
n
1) z
r,s
= (
n(pq)
1)z
r,s
= (
pq
1)t
0
.
Thus Kt
0
, . . . , t
n1

= S
q
. Then an argument as in the precedent lemma gives that S
j
S
k

= S
(n)
q
.
An analogous reasoning yields the following lemma.
Lemma 3.4. S
j
S
k

= S
(n)
j+k
as A
p
-modules for j + k < p and j ,= k.
Lemma 3.5. S
j
S
k

=

n1
l=0
P
l
= T
n
as A
p
-modules for j + k = p, where P
l
denotes the indecom-
posable projective ideal code corresponding to the simple ideal code X
l
in T
n
.
Proof. As in the previous results let us write S
j
S
k
= Kz
r,s
: 0 r, s n 1 for z
r,s
= v
r
v
s
and let us show the action of g on z
r,s
g z
r,s
=
(r+s)

(pj+pk)
z
r,s
=
(r+s)

p
z
r,s
=
(r+s+l)
z
r,s
for some l since 1 =
p
= (
n
)
p
= (
p
)
n
and thus
p
=
l
for some l.
Let us dene t
i,r
= x
i
z
r,0
, i = 0, . . . , n 1. The actions of g and x on these elements are
g t
i,r
= g (x
i
z
r,0
) =
i
x
i
g z
r,0
=
i
x
i

(r+l)
z
r,0
=
(i+r+l)
t
i,r
,
x t
i,r
= x (x
i
z
r,0
) = x
i+1
z
r,0
= t
i+1,r
, i = 0, . . . , n 2,
x t
n1,r
= x
n
z
r,0
= (g
n
1)z
r,0
= (
(r+l)n
1)z
r,0
= 0.
As in Lemma 3.2 we get that t
0,r
, t
1,r
, . . . , t
n1,r
is a linear independent set of vectors and from the
denition of P
r+l
we get that Kt
0,r
, t
1,r
, . . . , t
n1,r

= P
r+l
.
Now we observe that Kt
0,r
, t
1,r
, . . . , t
n1,r
+Kt
0,s
, t
1,s
, . . . , t
n1,s
is direct for r ,= s.
If Kt
0,r
, t
1,r
, . . . , t
n1,r
Kt
0,s
, t
1,s
, . . . , t
n1,s
= 0, then it contains a simple module L and since
both Kt
0,r
, t
1,r
, . . . , t
n1,r
and Kt
0,s
, t
1,s
, . . . , t
n1,s
contain only a simple module, respectively, S
r1
or S
s1
, we get that L

= S
r1

= S
s1
which is a contradiction. Thus,
n1

r=0
Kt
0,r
, t
1,r
, . . . , t
n1,r
S
j
S
k
.
But by comparing the dimensions we get that
S
j
S
k

=
n1

r=0
Kt
0,r
, t
1,r
, . . . , t
n1,r

=
n1

l=0
P
l+r

=
n1

l=0
P
l
= T
n
.
742 Garca-Rubira J M et al. Sci China Math April 2013 Vol. 56 No. 4
By the precedent lemmas we get the following result.
Theorem 3.6. The tensor product S
j
S
k
is isomorphic as an A
p
-module to
(i) S
(n)
q
if j + k
p
q with q ,= 0.
(ii)

n1
r=0
P
r
if j + k
p
0.
3.3 Product of an indecomposable Taft code with a simple in the matrix part
The following theorem characterizes the structure of a tensor product of an indecomposable Taft code
N
r,s
and a simple code S
j
in the matrix part of the decomposition of A
p
.
Theorem 3.7. N
r,s
S
j

= S
(ns)
j

= S
j
N
r,s
as A
p
-modules for every j = 1, . . . , p 1 and r, s =
0, . . . , n 1.
Proof. Let z
k,l
= u
k
v
l
for n s k n 1 and 0 l n 1. The actions of g and x on these
elements are
g z
k,l
= g u
k
g v
l
=
(r+k)
u
k

pj
v
l
=
(r+k+l)

pj
z
k,l
, n s k n 1 and 0 l n 1,
x z
k,l
= 1 u
k
x v
l
+ x u
k
g v
l
= u
k
v
l+1
+ u
k+1

pj
v
l
,
= z
k,l+1
+
l

pj
z
k+1,l
, n s k n 2 and 0 l n 2,
x z
n1,l
= 1 u
n1
x v
l
+ x u
n1
g v
l
= u
n1
v
l+1
+ 0
l

pj
v
l
= z
n1,l+1
, l = 0, . . . , n 2,
x z
n1,n1
= 1 u
n1
x v
n1
+ x u
n1
g v
n1
= u
n1
(
pj
1)v
0
+ 0
l

pj
v
n1
= (
pj
1)z
n1,0
.
Since is an primitive n-th root of unit we get that
(r+k+l)
= 1 if and only if r + k + l
n
0. Now
for xed r, the equation k + l
n
n r has a dierent solution for each k.
Let us consider now z
k,l
such that r + k + l
n
0 and dene t
i
= x
i
z
k,l
. Then
g t
i
= g (x
i
z
k,l
) =
i
x
i
g z
k,l
=
i

(k+r+l)

pj
x
i
z
k,l
=
i

pj
t
i
.
As in previous reasonings t
i
: i = 0, . . . , n 1 is a linear independent set. The action of x on these
elements is:
x t
i
= x x
i
z
k,l
= x
i+1
z
k,l
= t
i+1
, i = 0, . . . , n 2,
x t
n1
= x (x
n1
z
k,l
) = x
n
z
k,l
= (g
n
1) z
k,l
= (g
n
1) t
0
=
(k+r+l)
(
pj
)
n
t
0
t
0
= (
pj
1)t
0
,
since
n
= and
n
= 1.
In this way, for each of the s possible elections of z
k,l
we get that
Kz
k,l
, x z
k,l
, . . . , x
n1
z
k,l

= S
j
.
Thus N
r,s
S
j

= S
(ns)
j
.
An analogous reasoning gives that S
j
N
r,s

= S
(ns)
j
.
Garca-Rubira J M et al. Sci China Math April 2013 Vol. 56 No. 4 743
4 Some practical considerations
We recall from [4, Section 5] that indecomposable projective codes in the Taft Hopf algebra (those
appearing in Lemma 3.5) are a concatenation of cyclic codes and a generator matrix of the indecomposable
projective P
l
(l = 0, . . . , n 1) is
M
l
=
1
n

1 1 1 0 0 0 0 0 0
0 0 0 1
(l1)

(n1)(l1)
0 0 0
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
0 0 0 0 0 0 1
(ln)

(n1)(ln)

.
Simple codes in the matrix part may be also described easily and this is given in [5, Section IV]. A
generator matrix for a simple code S
j
in the matrix part is given, as above, by a block matrix,
N
j
=
1
pn

B
0
0 0 0
0 B
1
0 0
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
0 0 0 B
n1

,
where 0 denotes a block of pn zeros and the block B
m
, m = 0, . . . , n1, of length pn, is placed in position
m. This block is given by B
m
= (B
m0
, . . . , B
mp1
) with each B
ml
being the block of length n given by
B
ml
=
lj
(1,
pj
,
2(pj)
, . . . ,
(n1)(pj)
).
When considering linear codes, the tensor product of two linear codes can be viewed as a concatenation
of them (we are considering the usual linear structure) and, thus, the minimum distance of this new code
is the minimum of the corresponding two minimum distances. What we have considered through this
paper is the tensor product of two ideal codes and we have got a structure as ideal code on this new code
induced by the comultiplication of the considered Hopf algebra. Now we have got that this structure
of ideal code can be characterized depending on the case and, from Lemmas 3.2, 3.4 and 3.3 we get
codes that do not include the codes involved in the tensor product. Codes obtained in these lemmas are
concatenation of copies of one determined code in each case. The rst dierence with the linear case
is that the dimension is not the sum of the codes. Secondly, by Lemmas 3.2, 3.4 and 3.3 and from the
above matrix N
j
, we get that the tensor codes obtained in these lemmas have minimum distance pn,
analogously to the case of linear codes. However, Lemma 3.5 oers interesting consequences from the
Coding Theory and Ring Theory points of view respectively. From the Coding Theory point of view
we have to remark that dimensions behaves as in Lemmas 3.2, 3.4 and 3.3, but the minimum distance
decreases. From Lemma 3.5 and the matrix M
l
we get that the minimum distance of the corresponding
tensor product is n, being the minimum distance of the original codes appearing in the tensor product
equal to pn. From the Ring Theory point of view we get that the tensor product of two ideal codes as in
Lemma 3.5 is not semisimple.
Lastly, but not least important, we can observe that tensor products of indecomposable codes in the
Radford algebra is always a concatenation of cyclic codes, cf. [4] and [5]. Thus, although they do not
have a great behavior concerning the minimum distance, they are of easy and fast implementation by
blocks and suitable in situations where a limited number of errors can appear in every block, such as the
so-called byte error correcting codes (cf. [8] for examples).
Acknowledgements This work was supported by projects MTM2008-03339 from the Ministerio de Cienica e
In, P07-FQM03128 and FQM0211 from Junta de Andaluca and TEC2009-13763-C02-02.
744 Garca-Rubira J M et al. Sci China Math April 2013 Vol. 56 No. 4
References
1 Berman S D. On the theory of group codes (in Russian). Kibernetika, 1967, 1: 3139; Cybernetics, 1969, 1: 2539
2 Dasc alescu S, Nastasescu C, Raianu S. Hopf Algebras. An Introduction. New York: Marcel Dekker, 2001
3 Greferath M, Nechaev A, Wisbauer R. Finite quasi-Frobenius modules and linear codes. J Algebra Appl, 2004, 3:
247272
4 Cuadra J, Garca-Rubira J M, Lopez-Ramos J A. Determining all indecomposable codes over some Hopf algebras. J
Comput Appl Math, 2011, 235: 18331839
5 Cuadra J, Garca-Rubira J M, Lopez-Ramos J A. Codes as ideals over some pointed Hopf algebras. In: Proceedings
of the 19th International Symposium on Mathematical Theory of Networks and Systems. Budapest, 2010, 12871291
6 Montgomery S. Hopf Algebras and Their Actions on Rings. In: CBMS, vol. 82. Providence, RI: Amer Math Soc, 1982
7 Radford D E. On the coradical of a nite-dimensional Hopf algebra. Proc Amer Math Soc, 1975, 53: 915
8 Rao T R N, Fujiwara E. Error-Control Coding for Computer Systems. New Jersey: Prentice Hall, 1989
9 Wood J A. Duality for modules over nite rings and applications to coding theory. Amer J Math, 1999, 121: 555575
10 Wood J A. Code equivalence characterizes nite Frobenius rings. Proc Amer Math Soc, 2008, 136: 699706


Information for authors
SCIENCE CHINA Mathematics, a peer review mathe-
matical journal cosponsored by Chinese Academy of
Sciences and National Natural Science Foundation of
China, and published monthly in both print and electronic
forms by Science China Press and Springer, is committed
to publishing high-quality, original results in both basic
and applied research.
Categories of articles:
Reviews summarize representative results and achie-
vements in a particular topic or an area, comment on the
current state of research, and advise on the research
directions. The authors own opinion and related discus-
sion are requested.
Articles report on important original results in all areas
of mathematics.
Brief reports present short reports in a timely manner
of the latest important results.
Authors are recommended to use the online submis-
sion services. To submit a manuscript, please visit
http://mc03.manuscriptcentral.com/scmath, get an ac-
count, and follow the instructions to upload the text and
image/table files.
Authors should also submit such accompanying materials
as a short statement on the research background, ar-
ea/subarea and significance of the work, a brief introduc-
tion to the first and corresponding authors including their
mailing address, post code, telephone number, fax num-
ber, and email address. Authors may suggest several
referees (please supply full names, addresses, phone,
fax and email), and/or request the exclusion of specific
reviewers.
All submissions will be reviewed by referees selected by
the editorial board. The decision of acceptance or rejec-
tion of a manuscript is made by the editorial board based
on the referees reports. The entire review process may
take 90 to 180 days, and the editorial office will inform the
author of the decision as soon as the process is com-
pleted.
Authors should guarantee that their submitted manuscript
has not been published before, and has not been sub-
mitted elsewhere for print or electronic publication con-
sideration.
Submission of a manuscript is taken to imply that all the
named authors are aware that they are listed as
co-authors, and they have seen and agreed to the sub-
mitted version of the paper. No change in the order of
listed authors can be made without an agreement signed
by all the authors.
Once a manuscript is accepted, the authors should send
a copyright transfer form signed by all authors to Science
China Press. Authors of one published paper will be pre-
sented one sample copy. If offprints and more sample
copies are required, please contact the managing editor
and pay the extra fee. The full text in Chinese and
in English opens freely to the readers in China at
www.scichina.com, and the full text in English is available
to overseas readers at www. springerlink.com.
Subscription information
ISSN print edition: 1674-7283
ISSN electronic edition: 1869-1862
Subscription rates:
For information on subscription rates please contact:
Customer Service
China: sales@scichina.org
North and South America:
journals-ny@springer.com
Outside North and South America:
subscriptions@springer.com

Orders and inquiries:
China
Science China Press
16 Donghuangchenggen North Street,
Beijing 100717, China
Tel: 86-10-64034559 or 86-10-64034134
Fax: 86-10-64016350
Email: sales@scichina.org
North and South America
Springer New York, Inc.
Journal Fulfillment
P.O. Box 2485
Secaucus, NJ 07096, USA
Tel: 1-800-SPRINGER or 1-201-348-4033
Fax: 1-201-348-4505
Email: journals-ny@springer.com
Outside North and South America
Springer Distribution Center
Customer Service Journals
Haberstr. 7, 69126
Heidelberg, Germany
Tel: 49-6221-345-0, Fax: 49-6221-345-4229
Email: subscriptions@springer.com
Cancellations must be received by September 30 to take
effect at the end of the same year.
Changes of address: Allow for six weeks for all changes
to become effective. All communications should include
both old and new addresses (with postal codes) and
should be accompanied by a mailing label from a recent
issue. According to 4 Sect. 3 of the German Postal
Services Data Protection Regulations, if a subscribers
address changes, the German Federal Post Office can
inform the publisher of the new address even if the sub-
scriber has not submitted a formal application for mail to
be forwarded. Subscribers not in agreement with this
procedure may send a written complaint to Customer Ser-
vice Journals, Karin Tiks, within 14 days of publication of
this issue.
Microform editions are available from: ProQuest. Further
information available at http://www.il.proquest.com/uni.
Electronic edition:
An electronic version is available at springerlink.com.
Production:
Science China Press
16 Donghuangchenggen North Street, Beijing 100717,
China
Tel: 86-10-64034559 or 86-10-64034134
Fax: 86-10-64016350
Printed in the Peoples Republic of China
Jointly Published by
Science China Press and Springer
Mathematics
CONTENTS Vol. 56 No. 4 April 2013
Indexed by:
SCI-CD
MR
Z Math
MathSciNet
math.scichina.com www.springer.com/scp
Nonlinear instability for nonhomogeneous incompressible viscous fluids
..............................................................................................
665
JIANG Fei, JIANG Song & NI GuoXi
Existence and regularity of solutions to semi-linear Dirichlet problem of infinitely degenerate elliptic operators
with singular potential term
................................................................................................................................................................
687
CHEN Hua, LUO Peng & TIAN ShuYing
The boundedness for commutators of maximal hypersingular integrals with rough kernels
...................................................................
707
CHEN YanPing, DING Yong & LI Ran
Pointwise decaying rate of large perturbation around viscous shock for scalar viscous
conservation law
.................................................................................................................................................................................
729
DENG ShiJin & WANG WeiKe
Tensor products of ideal codes over Hopf algebras
.................................................................................................................................
737
GARCA-RUBIRA J. M. & LPEZ-RAMOS J. A.
A new proof for the correctness of the F5 algorithm
...............................................................................................................................
745
SUN Yao & WANG DingKang
Zygmund functions on the real line and quasiconformal deformations
...................................................................................................
757
SHEN YuLiang, LIU HongXia & WANG LiJun
Kato's inequality and Liouville theorems on locally finite graphs
..........................................................................................................
771
MA Li & WANG XiangYang
Necessary and sufficient conditions for Lipschitz ergodicity and generalized ergodicity
......................................................................
777
ZHENG ZuoHuan
Weak Orlicz spaces: Some basic properties and their applications to harmonic analysis
........................................................................
789
LIU PeiDe & WANG MaoFa
Similarity invariants of operators on a class of Gowers-Maurey spaces
..................................................................................................
803
LIN LiQiong, ZHONG HuaiJie & ZHANG YunNan
Von Neumann algebras generated by multiplication operators on the weighted Bergman space: A function-theory
view into operator theory
....................................................................................................................................................................
811
HUANG HanSong
Convergence of weighted averages of noncommutative martingales
.......................................................................................................
823
ZHANG Chao & HOU YouLiang
Operator-valued martingale transforms in rearrangement invariant spaces and applications
..................................................................
831
JIAO Yong, WU Lian & POPA Mihai
Some conditional results for conditionally strong mixing sequences of random variables
......................................................................
845
YUAN DeMei & LEI Lan
Learning rates of regularized regression on the unit sphere
.....................................................................................................................
861
CAO FeiLong, LIN ShaoBo, CHANG XiangYu & XU ZongBen
New semidefinite programming relaxations for box constrained quadratic program
..............................................................................
877
XIA Yong
Articles
Progress of Projects Supported by NSFC