Title :
Multi-bit flip type gradient descent bit flipping decoding using no thresholds
Author :
Haga, Rachel ; Usami, Shogo
Author_Institution :
Postgrad. Course in Sci. & Technol., Meijo Univ., Nagoya, Japan
Abstract :
Gradient Descent Bit Flipping (GDBF) decoding has one of the best decoding performances of the bit flip type decoding algorithms for low density parity check code. Multi-bit type GDBF decoding, however, achieves good error performance only with suitable thresholds. In this paper, we propose multi-bit type GDBF decoding without using thresholds. The proposed method estimates the number of bits to flip from the received soft sequence using the characteristic of the probability density distribution of the additive white Gaussian noise channel. The error performance of the proposed method is almost equal to that of the conventional method despite not using thresholds.
Keywords :
AWGN channels; decoding; gradient methods; probability; sequences; additive white Gaussian noise channel; gradient descent bit flipping decoding; multibit flip type; probability density distribution; soft sequence; Bit error rate; Computational modeling; Iterative decoding; Linear programming; Maximum likelihood decoding;
Conference_Titel :
Information Theory and its Applications (ISITA), 2012 International Symposium on
Conference_Location :
Honolulu, HI
Print_ISBN :
978-1-4673-2521-9