부산대학교 도서관

Byte correcting perfect codes are developed to correct burst errors within bytes. If a code is byte correcting code and we say that the code is -burst correcting, meaning that it corrects a single burst of length or less within a byte. A byte correcting code is such that if ; denote the set of syndromes obtained from the th-byte of the parity check matrix and ; denote the set of syndromes obtained from the th-byte of the parity check matrix then . In an code, if there are bytes of size , then . Byte correcting codes are preferred where information stored in all bytes are equally important. But there are cases where some parts of the message are more important than other parts of the message, for example, if we have to transmit a message on a border location ' Shift Battalion (Bn) from Location A to Location B' then, we will focus more on the information shift, location and location i.e., bytes containing these information will be more important than others. In this situation, it is needed that during transmission these bytes should have no possibility of error. In other words, these bytes should be protected absolutely against any error. Keeping this in mind, we study burst error correcting capabilities of byte oriented codes in terms of byte protection level of each byte. If there is a byte error pattern of length in the transmission then all those bytes of the received pattern will be decoded correctly whose burst protection level is or more even though the code word may be decoded wrongly. Taking the code length to be divided into bytes with burst protection level of the th-byte as ; ; , we construct linear codes that we call byte protecting burst (BPB) codes and investigate their byte protecting capabilities in this paper. [ABSTRACT FROM AUTHOR]