Re: Basic tools of encryption: Transposition and substitution?
Available news archives: comp.lang.tcl - comp.lang.python - comp.security.firewalls - sci.crypt - comp.lang.php - comp.lang.javascript
Google
 
Web news.hping.org


sci.crypt archive

Re: Basic tools of encryption: Transposition and substitution?

From: Unruh <unruh-spam@physics.ubc.ca>
Date: Tue Jan 03 2006 - 00:48:28 CET



Phil Carmody <thefatphil_demunged@yahoo.co.uk> writes:



>Doug <doug@null.com> writes:


>> Peter Fairbrother wrote:


>> 


>> > In lossless substitution the entropy remains unchanged. In lossless


>> > compression the entropy also remains unchanged - no entropy is lost - the


>> > message space shrinks.



In general the message expands. It is only for messages with a lot of


repetitions that the message space shrinks.



>> 


>> Hm. Not true:


>> 


>> Eg.


>> AABBBBAABBBB -> AA           (compressed) Code: A = AABBBB


>> AABBBBAABBBB -> BBCCCCBBCCCC (substitution)


>> 


>> H(AA) != H(BBCCCCBBCCCC)


>> H(x) = - Sum Pn * log(Pn)



>Take care. If you change the meaning of your tokens, you need 


>to change the model by which you are measuring entropy.


>In particular, your above example does not prove your 'Not 


>true' assertion.


Received on Tue Jan  3 03:42:46 2006