رسائل ماجيستير
English
2001
Modified algorithms for the decoding of hamming product codes
Maan A. Kousa, علي أحمد صالح الشيخي
كلية الهندسة-جامعة الملك فهد للبترول والمعادن · السعودية
Product codes attracted many researchers resulting in an explosive amount of literature. This is because product codes deal with the concept of constructing long powerful codes without increasing the complexity of the associated decoding, by employing shorter component codes. Also, product codes are effective in recovering both random and burst errors. Most proposed decoding algorithms of product code either do not exploit the full capability of the code or achieve high correction capability at the cost of increased complexity and delay. The objective of this thesis is to improve the performance of the product codes without a significant increase in decoding complexity. Specifically, the thesis introduces three modified algorithms. The modified hard decision decoder MHDD, algorithm develops a decoding algorithm that is capable of reaching the theoretical error correction capability of the product code without a significant increase in the decoding complexity. The extension of the MHDD to three dimensions is presented. The extended MHDD algorithm is also capable of reaching the theoretical error correction capability. The modified soft decision decoder, MSDD, algorithm investigates the performance of soft decision decoding for binary symbols to improve the performance by taking advantage of soft decoding in both dimensions. The modified erasure decoder, MED, algorithm investigates erasure decoding for binary symbols to improve the performance by erasing specific values instead of erasing whole rows or columns.