• Journal of Communications and Networks
• /
• v.17 no.4
• /
• pp.351-361
• /
• 2015

In this paper, a new decoding scheme is proposed to improve the error correcting performance of low-density parity-check (LDPC) codes in high signal-to-noise ratio (SNR) region by using post-processing. It behaves as follows: First, a conventional LDPC decoding is applied to received LDPC codewords one by one. Then, we count the number of word errors in a predetermined number of decoded codewords. If there is no word error, nothing needs to be done and we can move to the next group of codewords with no delay. Otherwise, we perform a proper post-processing which produces a new soft-valued codeword (this will be fully explained in the main body of this paper) and then apply the conventional LDPC decoding to it again to recover the unsuccessfully decoded codewords. For the proposed decoding scheme, we adopt a simple product code structure which contains LDPC codes and simple algebraic codes as its horizontal and vertical codes, respectively. The decoding capability of the proposed decoding scheme is defined and analyzed using the parity-check matrices of vertical codes and, especially, the combined-decodability is derived for the case of single parity-check (SPC) codes and Hamming codes used as vertical codes. It is also shown that the proposed decoding scheme achieves much better error correcting capability in high SNR region with little additional decoding complexity, compared with the conventional LDPC decoding scheme.

• Journal of the Korean Institute of Telematics and Electronics A
• /
• v.30A no.3
• /
• pp.16-33
• /
• 1993

New decoding algorithm of double-error-correction Reed-Solmon codes over GF(2$^{8}$) for optical compact disks is proposed and decoding algorithm of RS codes with triple-error-correcting capability is presented in this paper. First of all. efficient algorithms for estimating the number of errors in the received code words are presented. The most complex circuits in the RS decoder are parts for soving the error-location numbers from error-location polynomial, so the complexity of those circuits has a great influence on overall decoder complexity. One of the most known algorithm for searching the error-location number is Chien's method, in which all the elements of GF(2$^{m}$) are substituted into the error-location polynomial and the error-location number can be found as the elements satisfying the error-location polynomial. But Chien's scheme needs another 1 frame delay in the decoder, which reduces decoding speed as well as require more stroage circuits for the received ocode symbols. The ther is Polkinghorn method, in which the roots can be resolved directly by solving the error-location polynomial. Bur this method needs additional ROM (readonly memory) for storing tthe roots of the all possible coefficients of error-location polynomial or much more complex cicuit. Simple, efficient, and high speed method for solving the error-location number and decoding algorithm of double-error correction RS codes which reudce considerably the complexity of decoder are proposed by using Hilbert theorems in this paper. And the performance of the proposed decoding algorithm is compared with that of conventional decoding algorithms. As a result of comparison, the proposed decoding algorithm is superior to the conventional decoding algorithm with respect to decoding delay and decoder complexity. And decoding algorithm of RS codes with triple-error-correcting capability is presented, which is suitable for error-correction in digital audio tape, also.

• ETRI Journal
• /
• v.35 no.5
• /
• pp.767-774
• /
• 2013

In this paper, a novel parallel Viterbi decoding scheme is proposed to decrease the decoding latency and power consumption for the software-defined radio (SDR) system. It implements a divide-and-conquer approach by first dividing a block into a series of subblocks, then performing independent Viterbi decoding for each subsequence, and finally merging the surviving subpaths into the final path. Moreover, a network-on-chip-based SDR platform is used to evaluate the performance of the proposed parallel Viterbi decoding scheme. The experiment results show that our scheme can speed up the Viterbi decoding process without increasing the BER, and it performs better than the current state-of-the-art methods.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.22 no.2
• /
• pp.325-333
• /
• 1997

In fractral decoding procedure, the reconstructed image is obtained by iteratively applying the contractive transform to an arbitrary initial image. But this method is not suitable for the fast decoding because convergence speed depends on the selection of initial image. Therefore, the initial image to achieve fast decoding should be selected. In this paper, we propose an initial image estimation that can be applied to various decoding methods. The initial image similar to the original image is estimated by using only the compressed data so that the proposed method does not affect the compression ratio. From the simulation, the PSNR of the proposed initial image is 6dB higher han that of ones iterated output image of conventional decoding with Babaraimage. Computations in addition and multiplication are reduced about 96%. On the other hands, if we apply the proposed initial image to other decoding algorithms, the faster convergence speed is expected.

• The KIPS Transactions:PartC
• /
• v.10C no.1
• /
• pp.27-32
• /
• 2003

An efficient soft-decision decoding algorithm for binary block codes it proposed. The proposed soft-decision decoding algorithm is implemented by a series of hard-decision decoding process. By the hard-decision decoding result, the candidate codewords are efficiently searched for A new decoding method, which prevents the missing of the candidate codeword, is proposed. Also, the method fir reducing complexity is developed. This method removes the practical complexity increase caused by the improved algorithm. There facts are confirmed by the simulation results for binary (63, 36) BCH code.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.34 no.11C
• /
• pp.1025-1028
• /
• 2009

This paper presents upper bounds of Maximum Likelihood (ML) decoding performance of a few irregular LDPC codes using the simple bound and ML input output weight distributions and it is shown that contrary to general opinion that as block length becomes longer, BP decoding performance becomes simply closer to ML decoding performance, before peak degradation, as block length becomes longer, BP decoding performance falls behind ML decoding performance more and after peak degradation, general opinion holds.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.35 no.7C
• /
• pp.594-597
• /
• 2010

This paper estimates belief-propagation (BP) decoding performance of moderate-length irregular low-density parity-check (LDPC) codes with sphere bounds. We note that for moderate-length($10^3{\leq}N{\leq}4\times10^3$) irregular LDPC codes, BP decoding performance, which is much worse than maximum likelihood (ML) decoding performance, is well matched with one of loose upper bounds, i.e., sphere bounds. We introduce the sphere bounding technique for particular codes, not average bounds. The sphere bounding estimation technique is validated by simulation results. It is also shown that sphere bounds and BP decoding performance of irregular LDPC codes are very close at bit-error-rates (BERs) $P_b$ of practical importance($10^{-5}{\leq}P_b{\leq}10^{-4}$).

• Journal of Semiconductor Engineering
• /
• v.1 no.2
• /
• pp.74-80
• /
• 2020

Due to its superior error correcting performance with affordable hardware complexity, the Polar code becomes one of the most important error correction codes (ECCs) and now intensively examined to check its applicability in various fields. However, Successive Cancellation (SC) decoding that brings the advanced Successive Cancellation List (SCL) decoding suffers from the long latency due to the nature of serial processing limiting the practical implementation. To mitigate this problem, many decoding architectures, mainly divided into pruning and parallel decoding, are presented in previous manuscripts. In this paper, we compare the recent SC decoding architectures and analyze them using a tree structure.

• Proceedings of the IEEK Conference
• /
• 2003.07b
• /
• pp.803-806
• /
• 2003

본 논문에서는 traceback 동작 없이 decoding이 가능한 Modified Register Exchange 방식을 이용하여 이를 block decoding에 적용하는 비터비 decoding 방식을 제안하였다. Modified Register Exchange 방식을 block decoding에 적용함으로써 decision bit 들을 결정하기 위해 필요한 동작 사이클을 줄였고, block decoding을 사용하는 기존의 비터비 디코더보다 더 적은 latency 가지게 되었다. 뿐만 아니라, 메모리를 더 효율적으로 사용할 수 있으면서 하드웨어의 구현에 있어서도 복잡도가 더 감소하게 된다. 제안된 방식은 같은 하드웨어 복잡도로도 메모리의 감소 또는 latency 의 감소에 중점을 둔 설계가 가능하다.

• ETRI Journal
• /
• v.28 no.6
• /
• pp.790-792
• /
• 2006

The performance of a turbo decoder depends strongly on the number of iterations in its decoding process. It is necessary to stop the decoding process at an appropriate moment to alleviate the serious burden, in terms of both the computational speed and latency, part of which is associated with too many iterations. In this letter, we introduce a criterion for finding the opportune moment to stop the decoding process, called a hard decision aided criterion based on bit interleaved parity, which is known to have much simpler hardware logic, compared with other schemes, and does not lead to any significant performance degradation.