2.11 ‘RTP Statistics’ Worksheet
2.11.1 Interarrival Jitter
For the KPI definition, see IETF RFC 3550.
• Statistics tabulated as follows:
| Mean | Max | Min | Std Dev |
---|
Interarrival Jitter | 21.71 | 406.6213 | 0 | 5.02 |
• Same data displayed in bar chart binned according to this legend:
| % of Total |
---|
Range | Value |
---|
< 5 | 0.2 |
5 to 10 | 0.2 |
10 to 20 | 18.2 |
20 to 50 | 81.4 |
> 50 | 0.1 |
2.11.2 Packet Loss Ratio
KPI definition:
Number of lost (‘not used’) RTP packages divided by number of expected received RTP packets during the reporting period. Reported approximately once per second.
The expected number of received packets is calculated as the difference between the last received RTP sequence numbers in the current reporting period and the preceding reporting period.
A simulated jitter buffer is used, meaning that a RTP packet which has arrived out of sequence is still considered OK (not lost) if it is ‘caught’ by the jitter buffer, that is, if the packet sequence number is at least equal to the highest sequence number minus the jitter buffer size.
The jitter buffer size is configurable in TEMS Discovery (‘JitterBufferSize’; see chapter
3). This parameter is expressed as a number of RTP packets: for example, if JitterBufferSize is set to 5 and each RTP packet contains 20 ms of audio, the jitter buffer will span a time interval of 100 ms. Note that if JitterBufferSize is set very high, the packet loss will in practice be equal to network packet loss.
Report contents:
• Statistics tabulated as follows:
| Mean | Max | Min | Std Dev |
---|
Packet Loss Ratio | 1.84 | 96 | 0 | 5.92 |
• Same data displayed in bar chart binned according to this legend:
| % of Total |
---|
Range | Value |
---|
< 5 | 89.7 |
5 to 10 | 5.0 |
10 to 20 | 3.8 |
> 20 | 1.6 |