STG (13) 10 - Path Loss and Propagation Models in SEAMCAT
STG (13) 10 - Path Loss and Propagation Models in SEAMCAT
STG (13) 10 - Path Loss and Propagation Models in SEAMCAT
STG(13)10
ECC
CEPT
WGSE STG
33rd Meeting of SE STG
Copenhagen, Denmark, 6 8 February 2013
N
Password protection required? (Y/N)
Issue: Underestimation of the path loss for very short distances by the
Extended Hata SRD model
SE24 identified a significant underestimation of the path loss in in the Extended Hata SRD model
used in SEAMCAT for very short distances, e.g. below 5 m (see M68 SE24 minutes section 4.6). In this
document, this issue is addressed closer and a solution is suggested.
The discussion and tests in PT SE24 have shown that, in special cases, the path loss calculation in
SEAMCAT results in negative losses (see Figure 1), this seems not reasonable, because this result
would mean a gain instead of attenuation by the propagation model.
Therefore a propagation test was made with the input parameters presented in Table 1. The result is
presented in Figure 1. The result shows that a number of trails are resulting in negativ path loss.
1/6
Figure 1: Extract of the histogram of path loss (Extended Hata - SRD): below 7.5 dB path loss
The reason for this strange behaviour is that the median or mean path loss is computed correct.
However, the added normal distribution with a given standard deviation results in the spreading of
the path loss around the median or mean value of the path loss. It has to be highlighted that this
result is only producible if the propagation model has a high variation and the distance is low.
The standard deviation is predefined in the extended Hata model. For the case a relative high
standard deviation and short distances (e.g. up 200 m) it is likely to occur that the path loss is much
lower than free space. The SEAMCAT implementation of Extended Hata and Extended Hata-SRD
models available on the SEAMCAT online manual, states - in contradiction to the results - shown in
Figure 2 that:
When L is below the free space attenuation for the same distance, the free space
attenuation should be used instead.
Two examples of the path loss with two models and different standard deviations are given below for
illustration.
Table 2: Values for Path loss (Extended Hata) calculation for Figure 2
Extended Hata
Frequency 900 MHz
RX Height 1.5 m
TX Height 30 m
Distance 1m 200m
Environment Rural
2/6
Scenario Outdoor - Outdoor
Variation Active default: 5.6-12 dB (red curve)
Table 3: Values for Path loss (Free Space) calculation for Figure 2
Free Space
Frequency 900 MHz
RX Height 1.5 m
TX Height 30 m
Distance 1m 200m
Variation Active sigma = 3.5 (used in Extended Hata in
the first 40 m) (blue curve)
Figure 2: Results of path loss (Extended Hata and Free Space), between 80 to 140 m
In general, the propagation models could be used for very low distances, but the users have to have
in mind that the propagation models are fitting best the situation, if the model is used in the far field.
It is obvious that also in the illustrated examples the path loss can be significant underestimated.
One important assumption for all propagation models used in SEAMCAT is far field condition for the
antennas used: the path length d >> wave length
The boundary between the far field and near field can be roughly estimated with the assumption of a
plane wave front resulting in
2*D2 /
Where:
d = Distance
3/6
= Wavelength
D = maximum dimension of the antenna
Often used is the following far field condition for frequencies in the UHF range or higher.
d d = 11 *
d > 2*D2 /
d > 2*
1. Stick to the Free Space path loss if the calculated path loss including the random distribution
is less than the free space:
Advantage: Covers both cases described, no path loss < free space
Disadvantage: Deviates from equation in SEAMCAT manual
Illustration in Figure 3: In this option the blue line is the minimum allowed path loss.
4/6
Figure 3: Results of path loss (Extended Hata (red) and Free Space (blue)), between 1 to 200 m.
2. Stick to the Free Space path loss minus 3 dB if the calculated path loss including the random
distribution is less than the free space:
Advantage: Allows path loss less than free space, e.g. 2-way propagation
Disadvantage: Deviates from equation in SEAMCAT manual
3. Stick to the Free Space path loss with a normal distribution of standard deviation 3.5dB as
minimum pass loss for the Extended Hata model, in all environments and scenarios. This
sigma is used in the first 40m of the model (Free Space part).
Illustration in Figure 4: In this option the blue line is the minimum allowed path loss.
5/6
Figure 4: Results of path loss (Extended Hata (red) and Free Space (blue)), between 1 to 200 m.
6/6