User feedback – Nortel Networks NN43001-563 User Manual

Page 169

Advertising
background image

Post-installation network measurements

169

— number of cards and ports, IP Trunk 3.01 (and later) node and card

parameters

— fallback threshold level

— Codec image

— voice and fax payload

— voice and fax playout delay

— audio gain, echo cancellor tail delay size, Silence Suppression

threshold

— software version

User feedback

Qualitative feedback from users helps confirm if the theoretical QoS settings
match what end users perceive. The feedback can come from a Helpdesk
facility and must include information such as time of day, origination and
destination points, and a description of service degradation.

The fallback threshold algorithm requires a fixed IP Trunk 3.01 (and later)
system delay of 93 ms, which is based on default IP Trunk 3.01 (and later)
settings and its delay monitoring probe packets. The fallback mechanism
does not adjust when IP Trunk 3.01 (and later) parameters are modified
from their default values. Users can perceive a lower quality of service than
the QoS levels at the fallback thresholds in the following situations:

Delay variation in the intranet is significant. If the standard deviation of
one-way delay is comparable with the voice playout maximum delay, it
means that there is a population of packets that arrive too late to be
used by the IP Trunk 3.01 (and later) node in the playout process.

The jitter buffer is increased. In this case, the actual one-way delay is
greater than that estimated by the delay probe.

The codec is G.711A or G.711U. The voice packets formed by these
codecs are larger (120 to 280 bytes) than the delay probe packets (60
bytes). This means there is greater delay experienced per hop. If there
are low bandwidth links in the path, then the one-way delay is noticeably
higher both in terms of average and variation.

Nortel Communication Server 1000

IP Trunk Fundamentals

NN43001-563

01.01

Standard

Release 5.0

30 May 2007

Copyright © 2007, Nortel Networks

.

Advertising