Subscribe
Issue No.04 - April (1998 vol.47)
pp: 492-495
ABSTRACT
<p><b>Abstract</b>—A common method for constructing <it>t</it>-error correcting/all unidirectional error detecting (<it>t</it>-EC/AUED) codes is to append a tail to <it>t</it>-EC code, where the tail is a function of the weight of the codeword. Here, we present a method to reduce the weight distribution of <it>t</it>-EC codes by using a <it>DC</it><sub><it>m</it></sub>(<it>k</it> + <it>d</it>, <it>k</it>) code as the input of <it>t</it>-EC code where a <it>DC</it><sub><it>m</it></sub>(<it>k</it> + <it>d</it>, <it>k</it>) code is of length <it>k</it> + <it>d</it> and 2<super><it>k</it></super> codewords with the weight distribution between <tmath>$\lceil (k + d - m)/2 \rceil$</tmath> and <tmath>$\lceil (k + d + m)/2 \rceil.$</tmath> It is shown that our <it>t</it>-EC/AUED codes have the better information rates than the previously developed codes due to the smaller tails.</p>
INDEX TERMS
Descending tail matrix, unidirectional errors, balanced codes, DCm codes.
CITATION
Ching-Nung Yang, Chi-Sung Laih, "DCm Codes for Constructing t-EC/AUED Codes", IEEE Transactions on Computers, vol.47, no. 4, pp. 492-495, April 1998, doi:10.1109/12.675719