Search

Nonbinary LDPC codes demonstrated excellent performance. However, how to reduce the decoding complexity for such codes without performance degradation is still a problem to be solved. In this paper, the bit-flipping method is introduced into the standard Fourier transform decoding algorithm for decoding nonbinary LDPC codes. Simulation results show that the proposed method gains better BER performance with efficiently reduce of the decoding computation complexity in high Eb/N0 regions.

In this paper we propose algorithms which combine Belief-propagation (BP) decoding algorithm with genetic algorithm (GA). The main idea is to improve the performance of traditional BP decoding algorithm by efficiently using the belief of variable nodes passing to check nodes. By considering the belief as genes, the proposed algorithms mainly employ the mutation operation in GA, and change the genes? amplitudes (either increase or decrease) when updating the check nodes? information. First, we explore that the marriage of the GA and BP decoding algorithm is reasonable. Three algorithms are proposed accordingly. Proposed algorithm I mutates chosen genes by amplitude enhancement, while proposed algorithm II by amplitude reduction. Finally, proposed algorithm III is a mixture of the former two. From computer simulation results, we show that for short and middle length LDPC codes, our proposed algorithms can improve both BER (bit error rate) and FER (frame error rate) performance over traditional BP decoding algorithm, with slight modification of traditional BP decoding algorithm and little increase of computation complexity.

One of the obstacles in data analysis tasks is the variable quality of the data. We investigated ways to automatically deal with corruptions in the data. These include robust measures which avoid over fitting, interpolation-based imputation of missing values, and polishing by which the corrupted elements are fitted with more appropriate values. We applied such methods to a data set of vegetation indices and land cover type assembled from NASA's Moderate Resolution Imaging Spectroradiometer (MODIS) data collection. The experimental comparison suggested that straight-forward Interpolation, although a commonly used technique for filling in missing values, may not always yield the best results. Robust algorithms and polishing are both viable alternatives for dealing with partially corrupted data, with polishing performing slightly better in our experiments in terms of classification accuracy.