Propagation, Attenuation, and Noise

Propagation describes how a signal travels through a transmission medium (whether
metallic, radio, or optical). The signal energy will generally be confined within
bounded media (mainly cables), but radio signals can follow a variety of paths from
transmitter to receiver, including direct line of sight, reflected (from buildings or
terrain), or refracted (e.g., in layers of the atmosphere).

Attenuation describes the loss of energy as the signal travels through the medium,
resulting in reduced amplitude, while noise will be picked up from other sources of
electromagnetic energy, such as nearby cables or magnetic coils.
The noise levels picked up are usually very small, but the cumulative effect over
distance, together with the attenuation of the original signal, can quickly degrade
the channel to a point where it becomes unintelligible or large data errors occur.
High-bandwidth (usually high data rate) channels tend to degrade faster. Highfrequency
signals also tend to degrade relatively quickly.

Transmission systems are designed to minimize attenuation and provide good
immunity to noise. Fiber-optic cable is exceptionally good on both these counts
and has become the medium of choice for high data rate (high-bandwidth) channels,
especially in the core network.

Even for copper or radio systems, we can still mitigate the problems of attenuation
and noise on long transmission paths by careful system design, and also by
increasing transmitted power to compensate, or amplifying or regenerating the system
at appropriate points in the transmission link before it becomes unintelligible.

Amplifying an analog signal once it has attenuated and also picked up noise
would result in a larger-amplitude signal that also retains the amplified noise.
Whether this affects the user experience would depend on the proportion of noise
in the overall signal and the modulation scheme used.

As long as a digital signal can be regenerated before the noise makes it difficult
to recognize whether the bits are 1s or a 0s, the noise can be eliminated, and a clean
set of data can be retransmitted. Of course, some mistakes will inevitably occur
because of the random nature of the noise that is picked up, but these low levels
of bit errors can be eliminated or minimized by advanced digital processing techniques
in the receiver.

In general, the “noise immunity” that digital signals experience compared to
analog systems ultimately manifests itself in many ways, including clearer voice
channels, low error rate data channels, higher-capacity radio systems (for the same
infrastructure costs), or longer transmission paths.