Table of Contents
Advertisement
18.1.5 RTP
When you make a VoIP call using SIP or H.248, the RTP (Real time Transport Protocol) is
used to handle voice data transfer. See RFC 1889 for details on RTP.
18.1.6 Voice Coding
A codec (coder/decoder) codes analog voice signals into digital signals and decodes the digital
signals back into voice signals. The VOP supports the following codecs.
• G.711 is a Pulse Code Modulation (PCM) waveform codec. PCM measures analog signal
amplitudes at regular time intervals (sampling) and converts them into digital bits
(quantization). Quantization "reads" the analog signal and then "writes" it to the nearest
digital value. For this reason, a digital sample is usually slightly different from its analog
original (this difference is known as "quantization noise").
G.711 provides very good sound quality but requires 64kbps of bandwidth.
• G.723.1 uses Low-Delay Code-Excited Linear Prediction (LD-CELP) to code audio in 30-
millisecond frames. The standard supports two bitrates, 6.3 kbps and 5.3 kbps.
provides toll-quality sound and requires very little bandwidth.
• G.726 is an Adaptive Differential Pulse Code Modulation (ADPCM) waveform codec that
uses a lower bitrate than standard PCM conversion.
Differential (or Delta) PCM is similar to PCM, but encodes the audio signal based on the
difference between one sample and a prediction based on previous samples, rather than
encoding the sample's actual quantized value. Many thousands of samples are taken each
second, and the differences between consecutive samples are usually quite small, so this
saves space and reduces the bandwidth necessary.
However, DPCM produces a high quality signal (high signal-to-noise ratio or SNR) for
high difference signals (where the actual signal is very different from what was predicted)
but a poor quality signal (low SNR) for low difference signals (where the actual signal is
very similar to what was predicted). This is because the level of quantization noise is the
same at all signal levels. Adaptive DPCM solves this problem by adapting the difference
signal's level of quantization according to the audio signal's difference level. A low
difference signal is given a higher quantization level, increasing its signal-to-noise ratio.
This provides a similar sound quality at all signal levels.
G.726 operates at 16, 24, 32 or 40 kbps.
• G.729 is an Analysis-by-Synthesis (AbS) hybrid waveform codec. It uses a filter based on
information about how the human vocal tract produces sounds. The codec analyzes the
incoming voice signal and attempts to synthesize it using its list of voice elements. It tests
the synthesized signal against the original and, if it is acceptable, transmits details of the
voice elements it used to make the synthesis. Because the codec at the receiving end has
the same list, it can exactly recreate the synthesized audio signal.
G.729 provides good sound quality and reduces the required bandwidth to 8kbps.
18.1.7 PSTN Call Setup Signaling
PSTNs (Public Switched Telephone Networks) use DTMF or pulse dialing to set up telephone
calls.
3.
At the time of writing, the VOP supports the 5.3 kbps bitrate only.
MSC1000G/1024G/1224G Series User's Guide
Chapter 18 VoIP
3
G.723.1
551
Advertisement
Table of Contents
Troubleshooting
