Re: CCITT in C++ for AVR

A colleague here at CMU did some interesting work looking at the
robustness of CRCs and their implementation. One interesting artifact of his work is that some of the assumptions about failures do not hold in ultra-dependable systems. You might want to take a look at:

http://www.ece.cmu.edu/~koopman/pubs/paultisch05_dsn_crc_ultradependable.pdf

Perhaps this paper may be overkill for you, you sound like you just want an algorithm, but the first couple of sections may help you in your understanding of it.
John

PeteS wrote:

The CRC algorithm is almost trivial to implement. This search gave over
2 million hits (of which at least some appeared useful)
http://www.google.com/search?hl=en&lr=&q=crc+implementation+c&btnG=Search

I implemented CRC16 and CRC32 some years ago - took maybe an hour once
I had the algorithm in front of me. The polynomial is irrelevant to the
fundamental implementation.

It's possible there's a free implementation for AVR - see
http://www.avrfreaks.net/

Cheers

PeteS

.

Relevant Pages

  • Re: The CSI Challenge of snex
    ... Not being able to tell which one or both have deliberate artifact is ... pattern could always be present in an apparently random pattern. ... Again, if any bias can be found in form or function, the detection of ... ahead of time, THE ALGORITHM DOES NOT WORK, BY DEFINITION! ...
    (talk.origins)
  • Re: Embrace and Extend, Borland Style
    ... > cryptography to cover totally unrelated areas. ... Cryptography is about security too. ... robustness of the algorithm, not its secrecy that leads to a good algorithm. ...
    (borland.public.delphi.non-technical)
  • Re: convex hull implementation in 2D
    ... CGAL also adresses the robustness issues mentioned ... Is there a paper that compares different convex hull algorithms ... fastest convex hull algorithm for point sets < 100 points. ... algorithm) and get speed (no floating-point arithmetic), ...
    (comp.graphics.algorithms)
  • Re: The CSI Challenge of snex
    ... Not being able to tell which one or both have deliberate artifact is ... pattern could always be present in an apparently random pattern. ... Again, if any bias can be found in form or function, the detection of ... determined algorithm, the algorithm cannot detect artifact even if it ...
    (talk.origins)
  • Re: Kendall/Spearman rank correlation
    ... the accuracy of an approximate sorting algorithm. ... Either one would allow you to compare several ... a more persuasive argument about robustness. ...
    (sci.stat.math)