Abstract
Low-Density Parity-Check code was proposed in 1962, which is a kind of linear block code with great performance. The decoding algorithm of LDPC code is a kind of message-passing algorithm, and it’s proved that its performance is approaching the Shannon limited by puting capacity, LDPC code is ignored since it’s demonstrated until recent years.
Common decoding algorithms of LDPC code are hard-decision algorithm and soft-decision algorithm. The former one is relatively simple, it can decode faster, and it demands less on the hardware. While hard-decision algorithm is simple, its decoding performance is relatively low. Soft-decision algorithm has made a full use of the message received, so it’s plicated. In this way, soft-decision decodes slower, demand more on the hardware, and its performance is much better than hard-decision. This paper is mainly focused on the hard-decision decoding algorithm.
This paper firstly gives a research on the hard-decision bit-flipping algorithm proposed by Gallager, and then researches the two-bit bit-flipping algorithm, which is evolved from the bit-flipping algorithm. After that, this paper analyzed the advantage and disadvantage of two-bit bit-flipping algorithm, and concluded the reason why two-bit bit-flipping algorithm failed to decode. Through this conclusion, this paper then researched about the properties of trapping sets, and proposed a algorithm to detect whether or not the decoding process may enter a trapping set. Using the trapping sets detecting algorithm, we improved the two-bit bit-flipping algorithm.
Finally we give a simulation of the improved two-bit bit-flipping algorithm. Through the simulation, we can find that the decoding performance of improved two-bit bit-flipping algorithm is better than Gallager’s bit-flipping algorithm. Because of the trapping sets detecting algorithm and the pre-stop mechanism, the number of iterations is much more reduced. In this way, the performance of improved two
ldpc码多位比特翻转译码算法研究word论文 来自淘豆网www.taodocs.com转载请标明出处.