P1: IML
Morris WL040/Bidgoli-Vol III-Ch-53 September 15, 2003 12:23 Char Count= 0
THETELEPHONYNETWORK 651signal (whether analog or digitally encoded) to accurately
and timely propagate.
Wire transmission includes wire pairs, coaxial cable,
and even single wires (where the “return path” is the phys-
ical common ground). With wire transmission, the trans-
ducer varies the electrical pressure, or voltage, applied at
the transmitting end of the transmission link in propor-
tion to the transmitted signal, in effect pushing electrons
on one wire and pulling electrons on the other wire. The
electrons flow through this loop path to the node at the far
end of the transmission link. At the far end of the trans-
mission link, a transducer senses this electron movement
and pressure to recreate the transmitted signal, or some-
thing corresponding to it.
Fiber links involve the transmission of communica-
tions using a light signal through a very high-quality,
thin wire-like glass fiber. The signal is transmitted us-
ing a transducer that turns a light signal source (typi-
cally, a laser or light-emitting diode [LED]) at the infor-
mation source end of the transmission medium on and
then off or in varying intensities. Those variations are de-
signed to be some function of the transmitted information
signal. A “reverse” transducer at the far end node con-
nected to the fiber link senses these variations in light
intensity and ultimately deciphers the transmitted in-
formation.
Finally, wireless transmission links employ electro-
magnetic waves (such as radio and light waves) at the
transmitting end that are propagated through the open
atmosphere. The transmitted information is used to vary
some physical characteristic of the transmitted electro-
magnetic signal. A receiver-type transducer at the receiv-
ing end of the open-atmosphere medium detects these
variations in radio or light waves and, in turn, can
recreate the original signal and the information tran-
smitted.Transmission System Imperfections:
Noise, Loss, and Delay
Each transmission link in a network, and each node con-
necting those links, introduces some amount of noise,
loss, and delay into the transmitted signal. These physical
transmission imperfections affect both digitally and non-
digitally encoded signals. However, depending on whether
the signals transmitted are analog or digital, the trans-
mission channel’s physical limitations manifest them-
selves somewhat differently—as described later on in this
chapter.
Transmission noise introduced into the transmitted
signal during transmission causes the signal received at
the far end of a transmission link to vary from the origi-
nally transmitted source signal. Transmission noise is typ-
ically uncorrelated with the original transmission signal.
Where noise is present, detection and, therefore, recon-
struction of the original transmitted signal at the receiving
end is more difficult or impossible. Transmission noise,
when corrupting a digitally encoded signal, can result in
a receiver falsely identifying some of the 1s as 0s, and
some 0s as 1s.
The ability of a receiver to accurately detect the char-
acteristics of the original transmitted signal, and there-fore to accurately reproduce it, is directly a function of
the signal-to-noise ratio (S/N ratio) of the received sig-
nal. As the noise added by the transmission link becomes
very large in comparison to the power of the transmit-
ted signal, more errors in detection will occur. Boosting
the power of the original signal at the transmitting end
of the link can increase the signal-to-noise ratio, but this
is not always possible due to concerns about such prob-
lems as cross-talk between adjacent channels, and inter-
ference with other devices and systems. Special encod-
ing schemes can be used to allow the receiver to check
and/or correct for errors caused by random noise. The
particular schemes for doing this error detection and
correction are not important here, except to note that
they all require the “overhead” of transmitting additional
redundant information (and therefore additional bits)
along with the digitally encoded version of the original
signal.
There is one exception to the idea that noise is a bad
thing. Specifically, a limited amount of noise may some-
times be intentionallyaddedto a digitally transmitted
voice signal during quiet periods, when neither party is
speaking; otherwise the lack of noise could lead callers to
perceive that the line has been disconnected.
Transmission loss refers to the loss of signal power as
the original signal traverses a transmission link. Ampli-
fying all or selected frequencies of the received signal at
the receiving end of the link cannot always compensate
for signal loss. For example, in the presence of noise, if
there is significant overlap in the spectral power of the
noise signal and the original signal, receiver amplifica-
tion will tend to amplify the unwanted noise along with
the original transmitted signal. The result would be little
improvement in the receiver’s ability to discriminate be-
tween the now amplified original transmission signal and
amplified unwanted noise and thus its ability to accurately
detect the signal. The frequency range over which the loss
is relatively flat is known as the bandwidth.
Transmission delay (or latency) is a measure of how
much later a signal is received, as compared to when it
was transmitted. Delay can create two types of problems.
At the transmitted signal level, delay can vary with the
spectral range of the signal. For transmitted digital pulses,
this delay distortion can result in very distorted-looking
versions of the pulses arriving at the receiving end as com-
pared to the original signal. In turn, such delay distortion
can make accurate detection of the binary signal very diffi-
cult. If the delay distortion characteristics of a channel can
be determined in advance, they can be somewhat compen-
sated for at the signal detector. The second type of delay is
a delay in the time it takes for the information transmitted
(assuming it is accurately reproduced) to arrive at its ul-
timate destination. For voice conversations, this delay, if
long enough, can make voice communications almost im-
possible. When the round trip delay in a transmitted voice
signal reaches on the order of a half second, the receiving
and transmitting parties begin to notice this delay in two
ways. First, the parties at both ends begin to speak before
the other finishes speaking (because each party doesn’t
accurately know when the other has finished speaking).
Second, and more importantly, the parties at both ends
begin to hear their own voices echoed back to their ears.