.
1.4.2 Fast Fading vs Slow Fading
Similarly, based on the Doppler spread and the transmitted bandwidth (or equivalently
the coherence time and the symbol / frame duration), the fading channel is classified
as fusr fading or slow fading. A fading channel is classified as f i s t fading if the symbol
duration T, is larger than the coherence time T,. Otherwise, the fading channel is
called slow fading. Note that whether a transmitted signal experiences fast fading
or slow fading depends on both the channel parameters (such as the Doppler spread
or coherence time) and the transmitted signal parameters (such as the transmitted
bandwidth or the symbol duration). Hence, for slow fading channels, the channel
fading coefficients hi(t) remains to be quasi-static within a symbol duration T,.
Note that sometimes, we define slow fading to be the case when T, (coherence
time) is larger than Tf (frame duration) instead of symbol duration.
1.5 PRACTICAL CONSIDERATIONS
In this section, we briefly discuss the implications of path loss, shadowing and microscopic
fading components on system design. In fact, the design and performance of
communication systems depends heavily on the underlying channels. For instance,
path loss exponent determines how fast the received signal strength attenuates with
respect to distance separation between the transmitter and receiver. For a point to
point digital link, higher path loss exponents results in faster signal attenuation and
therefore is undesirable. This is because for the same transmitter and receiver design,
a higher path loss exponent results in shorter communication range (for the same
transmit power) or higher required transmit power (for the same distance) to overcome
the higher path loss. This holds in general for noise-limited systems. However, if we
consider multicell systems such as cellular networks like GSM, or CDMA systems,
higher path loss exponent results in more confined interference (as the interference
signal cannot propagate very far) and this translates into higher system capacity because
more aggressive resource reuse can be realized. In general, higher path loss
exponent is desirable for interference-limited systems. We will elaborate the cellular
system designs in Chapter 3.
Shadowing introduces randomness in the coverage of cellular systems. For instance,
if the standard derivation of the shadowing component is large, the average
received signal power at the cell edge will have large fluctuations. To achieve a certain
Quality of Service (QoS) such as 90% probability that the received signal strength at
the cell edge is above a target threshold, higher power or shorter cell radius is needed
to allow for some shadowing margins to satisfy the QoS target. Hence, shadowing
with larger standard derivation is undesirable.
Finally, the effects of microscopic fading have high impact on the physical layer
design of communication systems. For instance, by common sense, it is more challenging
to setup a wireless link to transmit high quality video then audio signals. This
is because for the same environment, transmitting a video signal generally involves
a higher transmit bandwidth, Wtr, relative to an audio signal. From Section 1.4.1, it
is very likely that the video signal will experience frequency selective fading while
the audio signal will experience frequency flat fading. As we elaborate in Chapter 4,
frequency selective fading will introduce intersymbol interference (ISI) and this induces
irreducible error floor. Hence, complex equalization at the receiver is needed.
On the other hand, for audio signal, since Wt, << B,, the signal will experience
flat fading only and no equalization at the receiver is needed. Hence, this results
in more simple design. Another common-sense example is that it is easier to setup
a wireless link for indoor environment than an outdoor environment to transmit the
same signal. This again can be analyzed by the frequency selective or frequency flat
fading channels. For the same transmit signal, the indoor environment in general
has a larger coherence bandwidth and hence, the number of resolvable multipaths
in indoor environment will be small. On the other hand, the coherence bandwidth
for outdoor environment in general will be small and this results in larger number of
resolvable multipaths (frequency selective fading). Hence, more complicated designs
are needed in the latter case.
To be Continue,,,,