IEEE TRANSACTIONS ON COMMUNICATIONS, VOL. 60, NO. 3, MARCH 2012 661
Transactions Papers
Nonbinary Cyclic LDPC Codes Derived from
Idempotents and Modular Golomb Rulers
Chao Chen, Baoming Bai, Member, IEEE, Zhuo Li, Xinquan Yang, and Li Li
Abstract—Based jointly on idempotents and modular Golomb
rulers, we construct a class of nonbinary cyclic low-density
parity-check (LDPC) codes. The defining parity-check matrix
is a sparse circulant, on which we put two constraints: 1)
the characteristic polynomial is an idempotent, 2) the nonzero
elements of the first row are located on a modular Golomb
ruler. We show that the second constraint forms a necessary
and sufficient condition for the Tanner graph to have no cycles
of length 4. The minimum distance of the code is proved equal
to the column weight of the parity-check matrix plus one. A
search algorithm is presented, with which we obtain some high
rate codes with large minimum distances. The issue of code
equivalence is also discussed. Simulation results show that the
obtained codes perform well under iterative decoding.
Index Terms—Nonbinary cyclic LDPC code, circulant, idem-
potent, modular Golomb ruler, code equivalence.
I. INTRODUCTION
C
YCLIC low-density parity-check (LDPC) codes are of
particular interest by virtue of their dual role [1]. On one
side, as cyclic codes, they can be simply encoded with shift
registers [2]. On the other side, as LDPC codes, they can pro-
vide very good error performance with a reasonable decoding
complexity based on message-passing algorithms [3], [4], [5].
For some cyclic LDPC codes, the parity-check matrix can be
transformed into quasi-cyclic form through row and column
permutations, which brings some decoding implementation
advantages [11]. These features make this class of codes very
attractive for practical applications. However, the available
constructions are not rich yet, especially for nonbinary codes.
In 2001, Kou, Lin and Fossorier constructed the first class of
cyclic LDPC codes based on finite geometries [7]. The parity-
check matrix is defined to be the incidence matrix of a finite
geometry, which can be arranged into a circulant or a column
Paper approved by T .-K. Truong, the Editor for Coding Theory and Tech-
niques of the IEEE Communications Society. Manuscript received February
27, 2011; revised July 26, 2011.
This work was supported in part by the NSFC and the 973 Program of China
under Grants 61101127, 2012CB316100, 60972046, 60902030, 61040005,
and 2010CB328300.
C. Chen and X. Yang are with the Science and Technology on Space Mi-
crowave Laboratory, China Academy of Space Technology (Xi’an), 710000,
China (e-mail: chenchaoxidian@gmail.com; yangxinquan@126.com).
B. Bai and Z. Li are with the State Ke y Lab of ISN, Xid-
ian University, Xi’an, 710071, China (e-mail: bmbai@mail.xidian.edu.cn;
lizhuo@xidian.edu.cn).
L. Li is with the China Academy of Space Technology (Xi’an), 710000,
China (e-mail: lili_504@126.com).
Digital Object Identifier 10.1109/TCOMM.2012.013112.110133
of circulants. The inherent geometric property ensures that
the Tanner graph has no cycles of length 4. As far as LDPC
codes are concerned, these codes have very large minimum
distances.
Subsequently, Shibuya and Sakaniwa introduced idempo-
tents to construct cyclic LDPC codes [8]. They pointed out that
if the number of cyclotomic cosets specifying an idempotent
is more than two, the search for such an idempotent with the
resulting Tanner graph having no cycles of length 4 become s
very difficult. Consequently, they focused mainly on the case
when an idempotent is specified b y at most two cyclotomic
cosets.
Operating in the Mattson-Solomon (MS) domain, Horan et
al. [10] systematically developed the idempotent based method
[8], [9]. They put forward three code design considerations:
1) the spareness of the parity-check matrix, 2) the code rate,
and 3) the minimum distance o f the code. By analyzing the
spectral properties o f these features, they were able to check
directly in the MS domain whether an idempotent gives a code
with desired properties.
Recently, Huang et al. [11] showed that, through row and
column permutations, a circulant can be transformed into
an array of sub-circulants; and an array of sub-circulants,
if satisfying certain constraints, can be transformed into a
circulant. Based on this and the sparse circulants available
from finite geometries, they constructed many new cyclic
LDPC codes. This work has greatly enlarged this class of
codes.
As for nonbinary cyclic LDPC codes, the existing construc-
tions are almost g eneralizations of the b inary codes mentioned
above. In [12], Zeng et al. constructed a class of nonbinary
cyclic LDPC codes based on finite geometries. The resulting
codes are either of short lengths or over large finite fields. In
[13], Tjhai et al. proposed a construction based on idempotents
and MS polynomials. With a search algorithm, they obtained
some short length codes over relatively small finite fields.
In this paper, we consider using idempotents and modular
Golomb rulers jointly to construct nonbinary cyclic LDPC
codes. We adopt a sparse circulant as the parity-check matrix.
With certain constraints on this circulant, the associated Tanner
graph is free of cycles of length 4 and the resultant code has
the minimum distance equal to the column weight plus one.
We then present a search algorithm. With it, we obtain some
short and moderate length codes over relatively small finite
fields, which have high rates and large minimum distances.
0090-6778/12$31.00
c
2012 IEEE