Re: [Hipsec] 5201bis -- Public key algorithms and key sizes

On 04/08/2010 04:59 PM, Robert Moskowitz wrote:
> It is our plan to add ECC as a PK algorithm in 5201-bis. My proposal 
> is to use:
>
> http://www.ietf.org/id/draft-mcgrew-fundamental-ecc-02.txt
>
> All of the math in this ID is pre-94. The ID makes no IPR claims but 
> the implication is there of unemcumbered ECC. Problem is there is no 
> standard libraries for the ID. 

I would like the HIP developers to look this ID over to see what it 
would take to implement. I am not aware of any code in standard 
libraries; I have asked Dave McGrew about this.

> First thing, is naming :) What might we call curves from this draft to 
> use in 5201bis?
>
> Aside from that P-256, P-384, and P-512 are recommended sizes based on 
> RFC 4753. Do we include P-384?
>
> Host-ID does not enumerate key size, nor does 5201 recommend a MUST 
> key size for interoperability (at least it is not clearly stated). I 
> contend that we recommend 2048 for RSA/DSA and for anything stronger, 
> ECC.
>
> Any point for RSA 1024? It does not match strength with the other 
> crypto components.
>
> Diffie-Hellman is all over the place with key sizes:
>
> 384-bit group 1
> OAKLEY well-known group 1 2
> 1536-bit MODP group 3
> 3072-bit MODP group 4
> 6144-bit MODP group 5
> 8192-bit MODP group 6
>
> I am not finding reference of DH MODP group size to strength. I would 
> think we would limit MODP key size to match RSA/DSA. For ECDH again 
> 256, 384 (maybe depending on question above) and 512.
>
> Comments?
>
>
> _______________________________________________
> Hipsec mailing list
> Hipsec@ietf.org
> https://www.ietf.org/mailman/listinfo/hipsec
>