Homophone encryption security & Vigenere
From: Markus Jansson (jansson_markus@ziplip.com)
Date: 02/19/03
- Next message: Cyber Vagrant: "Re: diehard and ent results quesion"
- Previous message: Bob Jenkins: "Re: ANNOUNCE: New "Leopard6" CSPRNG !"
- Next in thread: Mok-Kong Shen: "Re: Homophone encryption security & Vigenere"
- Reply: Mok-Kong Shen: "Re: Homophone encryption security & Vigenere"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ] [ attachment ]
From: Markus Jansson <jansson_markus@ziplip.com> Date: Wed, 19 Feb 2003 21:28:35 GMT
Im playing around with Cryptool 1.3 and wondering... How
secure are homophones? I remember that someone in this
groups stated that very large table of homophones would still be
very secure cipher. Is it really so and if it is, why arent
homophones used for encryption anymore (atleast in paper &
pencil ciphers)? How large should the table be for good level of
security? When using Cryptool and only encrypting letters, I end
up with only 4 decimal numbers...what if there where 1000
times more possible homophones?
>From what I can think of, homophones make actually very fast
paper & pencil cipher system. If you use numbers for instance,
you can arrange them in two tables: one that is arranged
according to the alphabets (to encrypt cleartext) and on that is
arranged according to the homophones (to decrypt ciphertext).
If you would have million homophones, they could easily fit into
two small books and arranged like a phonebook, some specific
range of alphabeths (or digits if its the one book used for
decrypting) in some pages and others in other pages etc. Using
random homophone for each time for the particular mark (from
the possible homophones available for that mark) would make
cracking it very hard I guess.
Anyway, what if the output of the homophone encryption (lets
say 7-digit numbers, each presenting one letter of plaintext, like
1111111 8374923 4617389) would be further encrypted using
Vigenere encryption? Adding and substracting numbers would
be very easy and very fast to do. The key(s) could be arranged
beforehand with the homophones codebook or agreed upon the
previous encrypted message (yes, I do know that it gives
possibility to known-plaintext-attack).
Anyway, am I making any sense here? Can homophone
encrypotion with very large keys be used securely even today
and wouldnt the addition of Vigenere make it even stronger?
Also, some other encryption methods could be used instead of
Vigenere after it has first been encrypted with homophone
encryption.
-- My computer security & privacy related homepage http://www.markusjansson.net PGPkeys fingerprints: 60D53AB1DA2509F7905803525EA153866E349FB2 1D993F2822460C9820D8F96369480E9A11A67E19
- Next message: Cyber Vagrant: "Re: diehard and ent results quesion"
- Previous message: Bob Jenkins: "Re: ANNOUNCE: New "Leopard6" CSPRNG !"
- Next in thread: Mok-Kong Shen: "Re: Homophone encryption security & Vigenere"
- Reply: Mok-Kong Shen: "Re: Homophone encryption security & Vigenere"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ] [ attachment ]