Page 207 - GuideFWA

3. ALLOWANCE FOR FADING

3.1. LMNQ VIEW

Normal planning rules for any wireless technology is based on a link budget formula such as:

Ptx = (I + N0) * CI / ( Lmean * Lshadow * Lmultipath )

where Ptx = Maximum Useable Transmit Power
I = Interference (typically from Internal System Sources)
Thermal Noise Power
N0 = minimum Carrier to Interference Ratio
CI = mean Path Loss from propagation attenuation
mean Shadow loss
Lmean = = Multipath loss
Lshadow =
Lmultipath

The Lmean “local mean” path loss is an attenuation factor which is averaged in time and over a
small area (e.g., 5-10 wavelengths ~1m @ 2GHz) to remove multipath variations. The path loss
is typically a function like aR-k where k is usually 2 for line-of-sight and between 2 and 6 for

non-line-of-sight, and R is the transmitter-to-receiver distance and typically is used to define the

maximum cell size.

Lshadow is typically due to static obstacles in the propagation path e.g., buildings, and the value
changes slowly due to mobility of the obstacle, transmitter or receiver.

Multipath fading is caused by the self-interference of reflected signal paths and are typically
rapidly varying in time (potentially <10 milliseconds between peaks) and in position. If there is
no line-of-sight path, the multipath fading distribution is typically Rayleigh, so the received
signal power is exponentially-distributed with a mean equal to the local mean power. If there is a
line-of-sight path plus scattered paths, a Rician model is often used. With Rayleigh fading, a
fade margin on the order of 10-17 dB must be incorporated into the link budget, with the exact
value depending on whether diversity is use, and the desired percentile point on the distribution.
For example, with no diversity, 17 dB corresponds to the 2% point, meaning that with a 17 dB
margin, there is a 2% probability that fading will drive the C/I or C/N below its threshold.

NOTE THAT USING THE SAME LINK BUDGET DESIGN GIVES A “GUARANTEE” OF
SUCCESSFUL INTERFERENCE DELIVERY i.e. a guarantee that the interference will be
higher than calculated for 98% of the time.

Inter-American Telecommunication Commission 193