> #" \pJava Excel API v2.6 Ba==h\:#8X@"1Arial1Arial1Arial1Arial + ) , * `&!DC,title[*]contributor[author]contributor[advisor]keywords[*]date[issued] publisher citationsidentifier[uri]identifier[doi]abstractrelation[journal]relation[volume]relation[no]relation[page]PHigh Performance and Low Complexity Decoding Algorithms for Nonbinary LDPC codesH02022. 2\ՑYPxhttp://hanyang.dcollection.net/common/orgView/200000591697;
https://repository.hanyang.ac.kr/handle/20.500.11754/168028;Although nonbinary low-density parity-check (LDPC) codes was introduced along
with binary LDPC codes, nonbinary LDPC codes did not receive much attention due
to its high decoding complexity. However, after several studies have shown that
the error correction performance of nonbinary LDPC codes exceeds that of binary
LDPC codes, numerous studies have been conducted to overcome their high
decoding complexity in recent decades. Based on these studies, many researchers
have tried to adopt nonbinary LDPC codes in communication system or data storage.
In this dissertation, we have studied both soft-decision and hard-decision based
decoding algorithms of nonbinary LDPC codes. For soft-decision based decoding
algorithm, we studied the methods that decrease the decoding complexity while
maintaining the error rate performance of extended min-sum (EMS) algorithm.
Unlike the original EMS algorithm, the proposed selection-based low-cost check
node operation decreases decoding complexity by using quick selection algorithm.
The experimental results show that the proposed check node operation is faster
and achieves even better error rate performance than the original EMS algorithm.
On the other hand, for hard-decision based decoding algorithm, we have studied
the methods that increase the error correction performance. The iterative pseudo
soft reliability-based majority logic decoding (IPSRB-MLGD) algorithm is modified
algorithm of iterative hard-reliability based MLGD (IHRB-MLGD) algorithm. The
IPSRB-MLGD algorithm utilizes the Hamming distance between hard-decision
symbol and other Galois field elements at initialization. The experimental result
show that the IPSRB-MLGD algorithm increases error rate performance and
reduces average iteration count. For further improvement of IHRB-MLGD algorithm,
we also proposed weighted IPSRB-MLGD (WIPSRB-MLGD) algorithm. In the
WIPSRB-MLGD algorithm, weighted voting method is applied in IPSRB-MLGD
algorithm and check node operation is modified to reduce the amount of
computation. As a result, The WIPSRB-MLGD algorithm achieves better error rate
performance with less amount of computation|Dt ĳ (Ӭ 8֔ t ĳ (Ӭ 8@ hخ <ǘ, 8 ĳ 买 D ȩD X. X̹ $X 1t t ĳ (Ӭ 8| 䲔 t lD t , \ D ٳH t 8X 8 LବX ĳ| t0 \ ι@ lt ɉ. t\ lD 0<\ Dt LDPC 8| pt0 ȥǥX ȩ¤ଐ XՔ l \ ɉ .
Y |8 Ƭ Dt LDPC 8X 8 LବX P x 0 8 Lବ 0 8 Lବ P \ l| ɉX. 0 8 LବD \ l\h Ƭ U֥ \͌i LବX $X 1@ Xt 8 1D tǔ )D lX. 0t U֥ \͌i Lବ , LବD XՔ 첬 |8 H 0 D xܴ Ű 0@ 5 LବD tǩX 8 ĳ| . H xܴ Ű 0@ 0tX U֥ \͌i Lବ Ŝ $X( 1D tt Ű ĳ h|ɔ D UxX.
\ո, 0 8 Lବ t Ƭ $X 1D tǔ ) t lX. Ƭ < 0tX \x 8 Lବx 0 ° LବD 0<\ Ǭ 0 ° 8 LବD HX. H\ Lବ x D ƌX t
p| 8 0T Ĭ X. , H\ Lବ@ 0t Lବ Dt $X( 1t `Oq\~Oq
!#
dMbP?_*+%" ,,??U
>@
!Root EntryWorkbook$SummaryInformation(DocumentSummaryInformation8