chapter one 23:23
Carrier-grade means high capacity, the capability to support hundreds of
thousands, perhaps millions, of subscribers and similar numbers of
VoIP is simply the transport of voice traffic using the Internet Protocol (IP),
hardly a surprising definition. However, it is important to note that VoIP does
not automatically imply voice over the Internet.
Here the IP is used because although it is less reliable but its fast as
compared to the tcp. And more reliable than the udp.
In case of voice transmission, every millisecond of data is not necessary.
For the traditional telephony circuit switching is used, but now for voip
IP is an attractive choice for voice transport for several reasons. These
Lower equipment cost
Lower operating expense
Integration of voice and data applications
Potentially lower bandwidth requirements
The widespread availability of IP
Circuit-switcCircuit-switched telephone networks transport voice at a rate of
64 Kbps. This is based on the Nyquist Theorem, which states that it is
necessary to sample an analog signal at twice the maximum frequency of
the signal in order to fully capture the signal. Typical human speech has a
maximum fre- quency of somewhat less than 4,000 Hz.
hed telephone networks transport voice at a rate of 64 Kbps. This is based
on the Nyquist Theorem, which states that it is necessary to sample an
analog signal at twice the maximum frequency of the signal in order to fully
capture the signal. Typical human speech has a maximum fre-quency of
somewhat less than 4,000 Hz. So 8000hz is required for capturing it fully.
For each sample 8 levels are used or 8 bits per sample are used which leads
to a bandwidth of 64kbps on both the sides for data transmission.
VoIP is not the only packet-based solution available to a com- mercial
carrier. In fact, Voice over Frame Relay (VoFR) and Voice over Asyn-
chronous Transfer Mode (VoATM) are powerful alternatives. One of the great
disadvantages of such solutions, however, is that they do not have the same
ubiquitous presence as IP . my notes 23:23
the lower the bandwidth, the lower the quality. my notes 23:23
Voice quality : Similar to the MOS value there exists a PSQM(Perceptual
speech quality measurement) this is an algorithm that attempts to faithfully
represent human judgement and perception processes. The result of PSQM
testing is a PSQM score that can be converted to an MOS value.
PSQM testing involves the algorithmic comparison between the output signal
of a communications system (such as a voice coder/decoder) and a known
input signal. The testing includes an algorithm that determines the combined
perceptual effects of a number of variables, such as the type of speaker
(male, female, or child), the loudness of the input signal, the delay, the
percentage of active/silent speech frames, clipping, and environmental
noise. Tests are conducted with a variety of different speech samples that
may have been produced artificially or by real speakers. If produced
artificially, then the speech samples should be generated according to ITU-T
According to the Nyquist theorem we can get back a signal only if we sample
it at a rate twice its maximum frequency.
After the sampling the samples has to be represented by some number of
bits. The number of bits are limited, e.g if we use 3 bits then 2^3 8 levels
are possible and if we need to represent 5.3 then we can only represent that
by 5. So there occurs this quantization error because now the 5.3 signal can
never be reproduced. So overcome this error we can use more number of
bits but that leads to the more utilization of bandwidth. So there are certain
methods by which the uneven quantization can be done and the signals can
be represented in a better way.
In this case the loud talkers sound bet