Wireshark jitter calculation

Question

I have a request regarding a method for calculating jitter in Wireshark.

Wireshark calculates jitter according to RFC3550 (RTP):

If Si is the RTP timestamp from packet i, and Ri is the arrival time in units of the RTP timestamp for packet i, then for two packets i and j D can be expressed as

D (i, j) = (Rj - Ri) - (Sj - Si) = (Rj - Sj) - (Ri - Si)

INTER-REGION MOVEMENT SHOULD be calculated continuously, since each data packet I is received from the source SSRC_n, using this difference D for this packet and the previous packet i-1 in the order of arrival (not necessarily in the sequence), according to the formula

J (i) = J (i-1) + (| D (i-1, i) | - J (i-1)) / 16

Now the absolute magnitude of jitter between arrivals is taken into account here. My request is why the absolute value was accepted, when the jitter can also be negative, and I think that if we take into account the negative jitter, we get a much actual value, and not the value that we currently take

In addition, when we plot the jitter distribution using the above method, it will not center around zero, since we made all the values ​​positive and this graph will not look realistic.

Can someone clarify my request?

Answer 1

Wikipedia has a good definition for jitter :

Jitter is an unwanted deviation from the true periodicity of the estimated periodic signal ...

A jitter value of zero means that the signal does not deviate from the expected value. As the variation increases (packets are compressed and scattered), the jitter increases in magnitude.

Clots and scattering are indeed the same effect; grouping in one place leads to spreading in another, therefore this "grouping and distribution" has no direction, just magnitude.

Hope this helps - this was the best explanation I could come up with.

Answer 2

See if this paper helps. Start reading from page 30 - it provides 4 methods. ftp://ftp.iol.unh.edu/pub/mplsServices/other/QoS_Testing_Methodology.pdf