Summary: | 碩士 === 中華大學 === 資訊工程學系碩士班 === 99 === Low-Density Parity-Check (LDPC) code is a powerful error correction code. Although first proposed in the 1960’s, it was not implemented due to its high hardware complexity. Recently, with the advance of VLSI technology, LDPC codes find extensively application in communication systems. Rahnavard et al. proposed how to achieve Unequal Error Protection (UEP) in LDPC code by appropriate code design. Their code design assigns different degrees to different groups of data bits. A dynamic scheduling decoding algorithm called Informed Dynamic Scheduling (IDS) was introduced to achieve configurable UEP for LDPC codes. The algorithm multiplies the residual of each group with different coefficients, trying to change the bit error rate (BER) of each group by altering their decoding schedules. Unfortunately, the attempt was not successful. Sometimes, the results even go in the opposite direction. Their strange results motivated our research. In this paper, we analyzed the coefficient method and found that the coefficient multiplication indeed changed the number of updates. However, it didn’t change the bit error rates. Why this phenomenon happen remains unanswered and it will need further studies.
|