I am analyzing some voice data. I have a data and each row represent a call scenario. Now each row has columns related to call.
The column I am interested in are SIP release cause, Jitter, MOS.
Here are some Release Causes:
'SIP: 487 Request Canceled', 'SIP: 200 OK', 'SIP: 480 Temporarily Unavailable', 'SIP: 504 Gateway Timeout', 'SIP: 408 Request Timeout', 'SIP: 484 Address Incomplete', 'SIP: 403 Forbidden', 'SIP: 404 Not Found', 'SIP: 486 Busy Here', 'SIP: 500 Server Internal Error', 'SIP: 503 Service Unavailable', 'SIP: 500 Service Unavailable', 'SIP: 580 Precondition Failure', 'SIP: 481 Call Leg/Transaction Does Not Exist', 'SIP: 603 Decline', 'SIP: 488 Not Acceptable Here'
Now I can see in my data for SIP OK , there are some jitter and MOS values which is correct since some RTP packets would be exchanged.
But for other release cause like SIP 487 Cancel, I could see some of my rows have jitter and MOS values which is strange since RTP packets are not expected here. Similarly with other 4XX and 5XX release causes.
Now I want to understand the cases (Release Causes) where there may be some RTP packet exchange which may lead to jitter.
Am I altogether getting the concept of jitter wrong. Can it occur without any RTP exchanges. Or are there any RTP flow for release cause other than SIP 200 OK.
200 OK is not the only SIP response to result in RTP packets being sent.
Early Media (RFC 3960) allows media (RTP) to be sent before the session is established with the 200 OK response.
Early media allows callers to hear particular ringback tones when a call is in the ringing state, and RVAs playing announcements like "Unable to reach that destination", without sending a 200 OK and flagging the call as answered.
In short, if you're looking at RTP packets or RTCP packets to calculate Jitter, then there's RTP packets being sent, the the cause is probably Early Media.