Re: [AVT] SRTP: question about MKI length

hi
   RFC 3711 assumes that the key management system will set the maximum 
length for the MKI.  The use of an MKI function is common in video 
broadcasting where a key gets rotated at rates that may be less than 
one second.  TV conditional access vendors operate proprietary systems 
that rotate the key according to application needs (there is really no 
cryptographic need to rotate a 128-bit AES counter-mode key until 2^64 
packets have been encrypted using it - a very long time).  The MKI was 
added for this application - and vendors in this industry use various 
sizes for the key index, particularly to convey a variety of key 
management information over a broadcast channel.

   In general, there is no need to use an MKI.  If there is, I would 
expect that a small, one-byte MKI would suffice to handle cases where 
key rotation might be useful.

Mark
On Jul 20, 2005, at 11:18 PM, Usha Sharma wrote:

> There is no description in RFC 3711 for upper limit of MKI length and 
> range of MKI value. SDP (draft-ietf-mmusic-sdescriptions-11.txt) 
> defines that MKI value is a positive integer and MKI length could be 
> up to 128 byte. Is it worthwhile to use such big MKI value for voice 
> applications, considering the bandwidth overhead introduced by it. 
> What would be the optimal value of MKI length for most applications?
> _______________________________________________
> Audio/Video Transport Working Group
> avt@ietf.org
> https://www1.ietf.org/mailman/listinfo/avt

_______________________________________________
Audio/Video Transport Working Group
avt@ietf.org
https://www1.ietf.org/mailman/listinfo/avt