North American Network Operators Group

Date Prev | Date Next | Date Index | Thread Index | Author Index | Historical

Re: estimating VoIP data traffic size from VoIP signaling traffic size ?

  • From: Bill Stewart
  • Date: Sun Oct 23 21:44:45 2005
  • Domainkey-signature: a=rsa-sha1; q=dns; c=nofws; s=beta;; h=received:message-id:date:from:to:subject:cc:in-reply-to:mime-version:content-type:content-transfer-encoding:content-disposition:references; b=KIK7aEwZYV5CLGb9fnNF2uWd/o7jepe7GIZSwepbsIFR9nBWIkcq3fIc80/bmUALpESeVby3uI0WDka8e2CEmWHNA77J6PX4Y0cPSx9TKFF1n3sObVjo+10wHJ7mhBAARKdGsC/+a9UsHbHjfjsc8a3IoP+OMwC9tzVPmSNKRmo=

Media traffic volumes are generally not visible, because they're from
endpoint to endpoint, so unless you've got really detailed monitoring
(which the original poster said they didn't), you're not going to see
traffic between two phones in the same building, or traffic between
buildings that don't have the call manager in them.  Obviously the
measurement problems are different for ISPs, enterprises with IP-PBXs,
and VOIP companies.

Also, the amount of media volume not only depends on the codec, but
also on the length of the call, while the signalling volume mainly
depends on the number of calls.  So if your customers are averaging 3
minute calls, that's a much different ratio than if they're doing
10-second credit card validation calls or one-minute voicemail pickups
or 60-minute teleconferences.  If you had enough measurement
capability to estimate this, you could use that directly instead of
guessing from signalling traffic, but otherwise you're just guessing.