摘要：In this paper, the traditional BP decoding algorithm of LDPC code is investigated in detail. The large computation load is a shortcoming of traditional BP decoding algorithm, and an improved BP decoding algorithm is then proposed to deal with the problem. In the iterative decoding process, the wrong bit information only needs to be updated for the decoding algorithm, which consequently improves the efficiency and reduces time-delay of decoding. According to the improved BP decoding algorithm, the high-speed parallel encoder is designed, which can be used in various distributed sensor network applications. Furthermore, the error correction performance of this high speed parallel encoder in the Gaussian channel is studied. The simulation is provided to illustrate that the improved high speed parallel encoder has lower computational complexity and higher decoding speed based on the premise of a little decoding performance loss.
