Question on determining key strength (Diffie-Hellman Key Exchange)

Greetings everyone!

Im working on my thesis which implements the use of the Diffie-Hellman key exchange. One problem that I encounter was how to assess and evaluate its strength given N-size of the public keys used. Does anyone know what is the recommended key size to achieve security with the Diffie-Hellman key exchange? And in what way was it determined?

Sincerely,

Paolo Ferrer

[407 byte] By [paoloferrera] at [2007-11-26 18:48:10]
# 1

Well, Diffie-Hellman is a key exchange protocol, not a cypher. If you mean RSA, then the recommended minimum is 2048 bits. This is determined by estimating the amount of time it would take to break a shorter key by brute force. 256 bits can be broken on a PC in hours. 512 can be broken on several hundred PCs over a couple of days (this is all very rough stuff). 1024 could theoretically be broken by a computer that might be built in the next decade or so in under a decade so - or something like that. So, 2048 is the rule of thumb - but it depends what you need it for. To send a secure message to your grandmother, it's unlikely the whole world will pool their resources to learn the text of your message in 10 years.

If, on the other hand, this is an email to your Justice department's Whitehouse liaison, you might want 4096 bits.

Look up "Diffie Hellman Key Exchange" and RSA on wikipedia.org for some good references.

TimSpearsa at 2007-7-9 6:22:10 > top of Java-index,Security,Cryptography...