[VoiceOps] Question about Packet Jitter

[anorexicpoodle]

Any time you are pulling jitter statistics from a central element like
an SBC using RTP stats you will only really have one side of the
picture, which is how it saw the packets arrive and how it passed them
on. If the far end can produce RTCP then the SBC can use RTCP to produce
stats for what the far end is reporting as well, but yes you are
correct, you would need the far ends opinion on the jitter score. 

On Thu, 2010-10-28 at 17:23 -0400, Alex Balashov wrote:

> On 10/28/2010 05:02 PM, anorexicpoodle wrote:
> 
> > If the SBC is holding all media, is there a reason it cannot simply
> > report what the jitter on the egress leg was and skip the derived
> > jitter equation? This is a simple enough trick for Acme, and I cannot
> > imagine most of the other vendors wouldn't have similar functionality.
> 
> That wouldn't yield the effective, de facto jitter (once network 
> conditions play out) on the receiving end of the egress leg, would it?
> 


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://puck.nether.net/pipermail/voiceops/attachments/20101028/f5fce2cd/attachment.html>