Section 16.2.1


DShambroom
01-27-2009, 06:01 PM
I believe the statement in the next-to-last paragraph on page 851:

"An (N,K) code can have d as large as N-K."

is incorrect. For example, a (15,11) code can have d=3, not 4.
A correct statement is:

"An (N,K) code can correct up to e bit errors, for the largest integer value of e that satisfies equation (16.2.2), with the equality relationship == replaced by the inequality <=. Such a code has d=2e+1."

Also, the last paragraph on page 852 is incorrect. Reed-Solomon codes have an alphabet of bytes, not bits, and an RS(28,24) code can correct considerably more that 2 bit errors.

Bill Press
01-27-2009, 07:27 PM
Thanks, Dave. Not sure where we got the statements in the book. Can you recommend a reference?
Cheers,
Bill P.

DShambroom
01-27-2009, 10:30 PM
Greetings Bill,

My observation about the statement on page 851 is based on understanding the math expressed in equation (16.2.2). As for page 852, based on what I do know about Reed-Solomon codes it's clear that the statement in the book can't be right, but I must admit that I do not have a definitive reference about the algorithmic details in my library.

N.B., I still consider NR3 an invaluable reference.

Best regards,
--Dave