| 研究課題/領域番号 |
16K06373
|
|---|
| 研究機関 | 同志社大学 |
|---|
研究代表者 |
程 俊 同志社大学, 理工学部, 教授 (00388042)
|
|---|
| 研究期間 (年度) |
2016-04-01 – 2019-03-31
|
| キーワード | spatially coupled code / irregular code / LDPC code / channel capacity |
|---|
| 研究実績の概要 |
We are progressing according to the plan in 2017, and achieved the following results.
We have proposed spatially-coupled irregular LDPC (low density parity check) codes by irregular superposition matrices. Starting with a binary base matrix and replacing non-zero entries in the base matrix by superposition matrices produces a parity-check matrix and its corresponding code ensemble. The proposed codes have two features: a) flexible code rates because the size of the superposition matrix is arbitrarily adjusted, and b) guaranteed iterative decoding performance because the weight distributions of the superposition matrix are optimized, even in a finite coupling width. The numerical analysis of density evolution and Monte-Carlo simulations show that our proposed code ensembles approach the Shannon limits at arbitrary code rates and outperform the conventional codes. Moreover, we analytically estimate the bit error rate (BER) and frame error rate (FER) of the proposed codes in error floor region. The numerical results show that the FER performances in the error floor regions of the proposed codes are superior to the conventional irregular LDPC code ensembles at almost identical belief-propagation thresholds.
|
| 現在までの達成度 (区分) |
現在までの達成度 (区分)
2: おおむね順調に進展している
理由
As summarize in the Research Achievements section, the research word is smoothly progressing according to the plan in 2017. The proposed spatially-coupled irregular LDPC codes have the feature of flexible code rates, and thus are available for multiple-access channel.
|
| 今後の研究の推進方策 |
Although the proposed three (two in 2016, and one in 2017) kinds of the spatially coupled codes have the decoding performance approaching the channel capacity in finite code length, we still attempt to find a kind of spatially coupled codes with short code length for the practical purpose. So, in 2018, we will slightly change our original research plan. We will continue to propose more powerful spatially coupled codes with short code length. One way of our attempting is to couple Polar codes, since Polar code works relatively well in short code length. After that we will apply our proposed codes to some wireless communication scenarios followed our original plan.
|
| 次年度使用額が生じた理由 |
（理由）
2017年度に予定していた成果発表を2018年11月の国際学会で行うことになったため．
（使用計画）
上記理由から，次年度使用額を2018年11月頃に国際会議参加費に支出します．
|
