This preview shows page 1. Sign up to view the full content.
Unformatted text preview: 14: Illustration of an eight by eight array code.
bits, located in the shaded region. The array is ﬁlled out to a total of 64 bits by the addition of a
row and column of parity bits, selected so that all rows and columns have even parity. The idea is
that all 64 bits should be stored or transmitted. During storage or transmission, one or more bits
might be changed due to a physical deviation. A bit that has been changed is said to be in error,
and the binary array with ones at the locations of the bits in error is called the error pattern. There
are 264 error patterns, including the pattern with no errors. When the data is needed, the reader
or receiver ﬁrst checks to see if all rows and columns have even parity. If yes, the data is deemed
to be correct. If no, the data is deemed to be corrupted, and we say the errors were detected.
Since there may be backup copies in storage, or the data can possibly be retransmitted, the eﬀect
of errors is not severe if they are detected. But the eﬀect of errors can be severe if they are n...
View Full Document
- Spring '08
- The Land