A Man Manual
A Man Manual
A Man Manual
WinProp
Antenna Pattern Editor
User Manual
Table of Contents
Mathematical Symbols.....................................................................................5
1
Introduction ...............................................................................................6
Isotropic Radiator.....................................................................................................7
2.2
Overview ...............................................................................................................83
4.2 2D
4.2.1
4.2.2
4.2.3
4.2.4
4.2.5
4.2.6
4.2.7
4.2.8
Mathematical Symbols
C
D
DdB
E(, )
G
GdB
H(, )
Pt
r
S
Si
3dB
3dB
Directional characteristic
Directivity (linear value)
Logarithmic directivity [dB]
Electrical field strength [V/m]
Gain of antenna (linear value)
Logarithmic gain of antenna [dB]
Magnetic field strength [A/m]
Radiated power [W]
Radius [m]
2
Power density [W/m ]
2
Power density of an isotropic radiator [W/m ]
Angle Phi [Degree] in horizontal plane (spherical coordinate system)
Half width in the horizontal antenna pattern [Rad]
Theta angle [Degree]
Half width in the vertical antenna pattern [Rad]
Increment in Phi [Degree]
Increment in Theta [Degree]
Equivalent radiation angle [Rad]
Efficiency of antenna
Introduction
Radio network planning tools rely on accurate wave propagation models to predict the path
loss between two arbitrary points. Besides the shielding of objects and multi path
propagation (both considered in the propagation models) the antenna pattern of the
antennas used for the communication link influence the actual path loss. Therefore the
antenna pattern must be described accurately within the radio network planning tool.
For this purpose AWE Communications developed the tool AMan (Antenna Manager) to
handle antenna patterns with a convenient Windows GUI. The most important features of
AMan are:
Graphical display of antenna patterns (horizontal or vertical plane)
3D Display of antenna patterns
Conversion of commercially available antenna file formats (*.msi, *.pln,.)
Conversion from 2x2D pattern (horizontal & vertical) to 3D pattern
Graphical editor to define manually vertical and horizontal antenna patterns (drawn
with mouse and scanned bitmap of pattern as background image)
Fig. 1-1: Vertical (left), horizontal (middle) and 3D pattern (right) of a sector antenna
Besides these important features AMan offers the optional module MASC (Multiple Antenna
Scenario Configuration) which is especially important if the influence of the local
environment and the mounting of the antenna should be considered and/or if multiple
antennas are combined to radiate the same signal and the user wants to see the radiation
pattern of the whole configuration to include this actual pattern in his radio network planning
process and tool. MASC allows the user to
Compute the resulting antenna pattern if different single antennas are combined to
a multiple-antenna configurations (incl. individual phase shifters and power
splitters)
Consider the influence of the mounting of the antenna (masts, arms, tubes,
radoms, etc.) on the actual antenna radiation pattern
As AMan is able to save the computed patterns in different file formats, it is not limited to
the WinProp Radio Network Planning suite. It can also be used for many radio network
planning tools as most of the tools are able to read the *.msi antenna pattern file format.
2.1
Isotropic Radiator
The ideal antenna operating without any losses and radiating the same power density in all
directions is called an isotropic radiator. This antenna is not available in real world but it is
used as a reference for all real-world antenna patterns.
The power density Si radiated from an isotropic radiator is homogenously distributed over a
spherical surface (with the radiator in the center of the spherical surface). With the power Pt0
fed to the antenna this leads to the following power density Si in a distance d:
Si
2.2
Pt 0
4 d
Commercial antennas do not radiate the same power density into all directions (in contrast
to the isotropic radiator presented in section 2.1). There are always some directions with
higher power density and others with smaller power density. The direction with the highest
power density is called main direction.
The antenna pattern describes the dependency of the radiation on the direction. It shows
amplitude, phase and polarization assuming far field conditions (i.e. far away from the
antenna itself). Spherical coordinates and are generally used to describe the antenna
pattern. As the distance is not relevant in the far field, the antenna pattern itself is a
function of the angles and
The pattern of real antennas depend additionally on the frequency. So for different
frequencies different patterns must be used.
Very often the radiation of the antenna is only measured in the horizontal and in the vertical
plane. This reduces the effort and describes the antenna for most applications. Only very
accurate wave propagation models (like ray-optical models) will benefit from 3D antenna
patterns.
Horizontal Pattern:
Vertical Pattern:
3D Representation:
Vertical Pattern:
Horizontal Pattern:
3D Representation:
As the power density is proportional to the electric field strength, very often the electric field
strength E(,) is used to describe the antenna pattern. Usually the pattern is normalized on
its maximum value, e.g.:
CE ,
E ,
E , max
Since, in the general case, an elliptical polarization is present in the far field, both antenna
patterns of the orthogonal polarization, C1(,) and C2(), are needed for the description
of the directional characteristics of the antenna. C1(,) and C2() are generally
normalized on the maximum value Emax of the highest component:
C1
E1 ,
,
Emax
Cges
C2
E2 ,
Emax
C12 C22
2
1
C22 max
8
2.3
2.3.1
If the main direction of radiation is known, the directivity D can be calculated for each
antenna pattern with the following equation:
S
D max
Si r const , P const
t
Smax indicates the maximum radiated power density and Si indicates the power density
4
2
C , sin
2
d d
0 0
If the antenna pattern consists of a discrete number of values, the integrals in the equations
must be substituted with a summation with increments and :
2 / /
C i ,
2
i 1
j 1
j sin j
Very often the directivity of an antenna is given in dB. The logarithmic value DdB in dB can be
obtained from the linear value D of the directivity with the following equation:
DdB 10 log D
2.3.2
Gain
In contrast to the directivity (see section 2.3.1), the gain G of an antenna considers
additionally the efficiency of the antanna, i.e. not the radiated power density is compared
to the isotropical radiator but the power fed to the antenna. So losses in the antenna are
included in the value of the gain but not in the value of the directivity:
G=*D
AWE Communications GmbH
10
2.3.3
Characteristic values
Characteristic Values:
20 log(b)
20 log(c)
2 M
2 S
2.3.4
Half widths
The characteristic values given in section 2.3.3 are applied to the vertical and the horizontal
pattern. In case of vertical pattern the variable is used and in the horizontal pattern the
variable is used to describe the angles. The 3dB angles (see section 2.3.3) are also called
half widths.
Directivity is indicated in dB and the lobe width 3dB values are more easily legible in degrees
in a directivity pattern. It results in the following adapted dimensional equation (angles
values are in degree):
41253
M | M |
11
2.3.5
e is the equivalent angle which is calculated from the both sides of the half-values of the
angles, M and M, of the two planes.
e M M
The radiated power density within the equivalent angle is assumed to be equal to the max.
radiated power density (in main direction) and outside the equivalent angle no power is
emitted.
2.3.6
4
e
Effective area
Gain and directivity of an antenna in transmitter or receiver operation are equal (reciprocity).
Received power Pr is obtained from power density S (i.e. electromagnetic field).
Effective area of an antenna Ae is defined according to:
Pr = Ae * S
Isotropic antenna:
Aei = 2 / 4
Ae = D * Aei
12
2.3.7
13
Dipole
Dipole
Dipole
Dipole
Dipole
Dipole
Dipole
Dipole
Fig. 2-7: Horizontal pattern of dipole antennas (depending on the length of the dipole)
14
2.4
2.4.1
An antenna radiates in all directions. Therefore 3D patterns are required to describe the gain
individually for each direction. For some applications the radiation in the horizontal plane and
the radiation in the vertical plane is sufficient. These patterns are called 2x2D patterns.
The following two figures show the 3D pattern and the 2x2D patterns of a half wave dipole.
Hertz Dipole
Halfwave Dipole
Fig. 2-9: Horizontal pattern (left) and vertical pattern (right) of a halfwave dipole
It is obvious that accurate computations are only possible with 3D patterns. AMan includes
therefore a module to convert 2x2D patterns to 3D patterns (see section 2.5).
Antenna patterns can be defined in any coordinate system. Spherical coordinate systems are
most commonly used to define antenna patterns.
15
2.4.2
3D Antenna Patterns
3D patterns can be defined in any coordinate system. Spherical coordinate systems are most
commonly used to define antenna patterns.
Figure 2-10 shows the coordinate system used in WinProp for the 3D display of antenna
patterns. The main direction of radiation is = 0 and = 0.
Figures 2-11 to 2-13 show the 3D pattern as well as the vertical and horizontal plane of a
Kathrein K739856 antenna together with the coordinate systems used for display in AMan.
16
17
2.4.3
The radiation patterns in the horizontal and in the vertical plane are available for nearly all
commercial antennas. Full 3D patterns are not always available as they are more
complicated to be measured.
18
Fig. 2-15: Coordinate system of horizontal pattern (if viewed from top)
19
2.5
The 3D pattern of an antenna can be computed based on the pattern of the vertical and
horizontal plane. AMan offers four different conversion algorithms. The user can decide
which algorithm should be used.
2.5.1
The most simple algorithm searches for the two angles and the gain Gtilt in the vertical
() and the gain Grot in the horizontal () plane.
The arithmetic mean of Grot (horizontal pattern) and Gtilt (vertical pattern) represents the
actual gain G (, ).
G ,
Grot ( ) Gtilt ( )
2
Patterns generated with this algorithm are not very accurate. Therefore this algorithm is not
recommended (it is only implemented for comparison with other tools).
2.5.2
As defined in figure 2-17, four gain values are determined in the horizontal and vertical
patterns depending on the two angles and .
The gain values are weighted with their angle distances:
G ,
GV 2 GV G H 2 GV 0
2
Z
G
Theta
Ph
i
X
G
-Y
Fig. 2-17: Angles and gains used for the bilinear interpolation
AWE Communications GmbH
20
2.5.3
This algorithm is almost identical to the algorithm presented in section 2.5.1. The gains and
angle distances are also read from the vertical and horizontal pattern and are weighted
according to their distances.
In contrast to algorithm 2, the vertical angles are additionally weighted with the Factor (1-sin
). Therefore the gain values read from the vertical pattern are no longer relevant in the
horizontal plane (i.e. for = 90) and the horizontal pattern is therefore more accurate.
The mathematical equation for the computation of the weighted bilinear interpolation is:
G
1
G ,
12
1 2
1G 2 2G1
2
2
1 2
1 2
1 2 1 2 2 1 2 1 2 2
1 2
1 2
2G 1
P(theta, phi)
Gtilt2
d til
t2
Grot2
d ro
t2
Grot1
d tilt1
drot1
Ph
i
Gtilt1
-Y
Fig. 2-18: Angles and gains used for the weighted bilinear interpolation
G ,
dtilt1 Gtilt 2 dtilt 2 Gtilt1 drot1 1 sin Grot 2 drot 2 1 sin Grot1
dtilt1 dtilt 2 drot1 1 sin drot 2 1 sin
The weighted bilinear interpolation leads to more accurate results compared to algorithm 2.
The mean error is approx. 1.3 dB and the standard deviation is 0.6 dB for a /2-Dipole.
This algorithm (together with algorithm 2) was published by F. Gil, A. Claro, J. Ferriera, C.
Pardelinha, and L. Correia, with the title A 3D Interpolation Method for Base-Station-Antenna
Radiation Patterns in IEEE Antennas and Propagation Magazine, Vol. 43, No.2, April 2001.
The most accurate results can be obtained with this algorithm, if the main radiation of the
antenna is in the horizontal plane.
21
2.5.4
If the main radiation is not in the horizontal plane (e.g. if the antenna has an electrical or
mechanical downtilt), the pattern computed with algorithms 2 and 3 becomes less accurate.
Especially in these cases the HPI algorithm should be used.
The HPI algorithm takes the gain of the horizontal pattern GH as a basis and considers a
correction term for the influence of the vertical pattern GV . Therefore the gains GH
and GV in the horizontal and vertical pattern, respectively, are taken and processed by
using the following equation:
G H 0 GV G H GV
G , G H
Hereby it is assumed that the horizontal and vertical patterns are two sections of the 3D
antenna pattern. This means that the two following conditions are fulfilled:
G H 0 GV 0 and G H GV in the case without electrical tilt
G H 0 GV and G H GV in the case with electrical tilt
2.5.5
An alternative algorithm for the extrapolation of 3D antenna radiation patterns based on the
given horizontal and vertical patterns is the exponential interpolation. In this method the
horizontal diagram is corrected according to the difference between the front-to-back ratio at
elevation angle 0 and the front-to-back ratio at the elevation angle theta by using the
following equation:
GV GV
G , GV
G H GVMax with 360
GV 270 GV 90
This exponential interpolation provides accurate predictions especially for antennas with
electrical downtilt.
2.5.6
In this section the four presented interpolation algorithms are compared on one hand
visually when looking on the resulting 3D patterns and on the other hand analytically when
using the resulting 3D patterns for the path loss prediction. For the comparison three
different antennas from Kathrein are going to be evaluated with different values concerning
the horizontal and vertical half power beam width (HPBW):
Name
Type
Frequency
HPBW (H)
HPBW (V)
Max. gain
Elect. tilt
741 794
X-Pol
2110 MHz
65
18.5 dBi
741 984
X-Pol
1855 MHz
87
26
11.5 dBi
742 211
X-Pol
1855 MHz
67
14
15.5 dBi
Fig. 2-19: Antennas (and their characteristic values) taken into account for the comparison
As reference for the comparisons always the measured 3D antenna pattern is available (in all
three examples). These measured 3D patterns are provided by the manufacturer of the
antennas (Kathrein).
AWE Communications GmbH
22
In the following an individual subsection is dedicated to each antenna, comparing the results
of the different interpolation algorithms (on one hand by the resulting 3D antenna patterns
and on the other hand by predictions of the received power when using the corresponding
interpolated 3D antenna).
2.5.6.1
The antenna 741 794 is a sector antenna with 65 horizontal and 7 vertical half power
beam width and 2 electrical downtilt. The horizontal and vertical patterns of this antenna
are given in figure 2-20 while the real, i.e. measured, 3D pattern is visualized in figure 2-21.
The sketch of the 3D pattern is taken from a side view, i.e. the main lobe of the antenna
characteristic is oriented towards the right. The views in figure 2-22 represent the
interpolated 3D patterns for the four different interpolation algorithms. Similar to the
measured 3D antenna pattern also these figures show the side view of the antenna with the
main lobe oriented towards the right.
Fig. 2-20: 2D antenna patterns in horizontal and vertical plane (logarithmic scale)
23
Fig. 2-22: Interpol. 3D patterns by AM (top left), BI (top right), WBI (bottom left), HPI (bottom right)
AWE Communications GmbH
24
Fig. 2-23: Prediction of received power if using the measured 3D antenna pattern (Tx height 20 m)
Fig. 2-24: Prediction of received power using the interpolated 3D pattern (AM algorithm)
25
Fig. 2-25: Prediction of received power using the interpolated 3D pattern (BI algorithm)
Fig. 2-26: Prediction of received power if using the interpolated 3D pattern (WBI algorithm)
26
Fig. 2-27: Prediction of received power using the interpolated 3D pattern (HPI algorithm)
The plots of the predicted received power for the transmitter height of 20m show the
influence of the antenna pattern on the computation of the wave propagation. At the first
view all the figures look very similar, however after detailed analysis there are some
differences visible.
Comparison
Measured 3D
Pattern AM algorithm
BI algorithm
WBI algorithm
HPI algorithm
Polarization +45
Mean value [dB]
Std. Dev. [dB]
Polarization -45
Mean value [dB]
Std. Dev. [dB]
0.57
2.41
0.07
1.81
-0.32
2.14
-1.03
2.41
-0.77
2.10
-1.28
2.00
0.17
2.00
-0.20
1.42
The numerical evaluation of these differences is given in the table above. According to this
evaluation the HPI has the best performance, i.e. the smallest error with respect to the
measured 3D antenna pattern. However all the different interpolation algorithms are in the
same range (at least concerning the standard deviation).
27
Fig. 2-29: Prediction of received power using the measured 3D antenna pattern (Tx height 10 m)
Fig. 2-30: Prediction of received power using the interpolated 3D pattern (AM algorithm)
28
Fig. 2-31: Prediction of received power using the interpolated 3D pattern (BI algorithm)
Fig. 2-32: Prediction of received power using the interpolated 3D pattern (WBI algorithm)
29
Fig. 2-33: Prediction of received power using the interpolated 3D pattern (HPI algorithm)
The plots of the predicted received power for the transmitter height of 10m show the
influence of the antenna pattern on the wave propagation computation. Also for this reduced
transmitter height all the figures look very similar, however after detailed analysis there are
some differences visible.
Comparison
Measured 3D
Pattern AM algorithm
BI algorithm
WBI algorithm
HPI algorithm
Polarization +45
Mean value [dB]
Std. Dev. [dB]
Polarization -45
Mean value [dB]
Std. Dev. [dB]
0.19
1.58
0.07
0.78
-0.72
1.71
-1.43
2.21
-1.68
1.84
-2.06
1.85
-0.23
1.41
-0.34
1.11
Fig. 2-34: Numerical evaluation of the power predictions (measured 3D interpolated 3D) for Tx height 10m
The numerical evaluation of these differences is listed in the table above. According to this
evaluation the AM algorithm has the best performance, i.e. the smallest error with respect to
the measured 3D antenna pattern.
However, the HPI algorithm achieves nearly the same performance.
30
2.5.6.2
The antenna 741 984 is a sector antenna with 87 horizontal and 26 vertical half power
beam width and no electrical downtilt. The horizontal and vertical patterns of this antenna
are given in figure 2-35 while the measured, 3D pattern is visualized in figure 2-36. The
sketch of the 3D pattern is taken from a side view, i.e. the main lobe of the antenna
characteristic is oriented towards the right.
The views in figure 2-37. represent the interpolated 3D patterns for the four different
interpolation algorithms. Similar to the measured 3D antenna pattern also these figures show
the side view of the antenna with the main lobe oriented towards the right.
Fig. 2-35: 2D antenna patterns in horizontal and vertical plane (logarithmic scale)
31
Interpolated 3D Patterns
Fig. 2-37: Interpol. 3D patterns by AM (top left), BI (top right), WBI (bottom left), HPI (bottom right)
32
Fig. 2-38: Prediction of received power using the measured 3D antenna pattern (Tx height 20 m)
Fig. 2-39: Prediction of received power using the interpolated 3D pattern (AM algorithm)
33
Fig. 2-40: Prediction of received power using the interpolated 3D pattern (BI algorithm)
Fig. 2-41: Prediction of received power using the interpolated 3D pattern (WBI algorithm)
34
Fig. 2-42: Prediction of received power using the interpolated 3D pattern (HPI algorithm)
The plots of the predicted received power for the transmitter height of 20m show the
influence of the antenna pattern on the wave propagation computation. When comparing the
different power plots there are some differences visible concerning the main lobe and also
the side lobes.
Comparison
Measured
Pattern AM algorithm
Polarization +45
Mean value [dB]
Std. Dev. [dB]
BI algorithm
WBI algorithm
HPI algorithm
Fig. 2-43:
Polarization -45
Mean value [dB]
Std. Dev. [dB]
0.07
1.07
0.21
1.09
2.18
1.06
1.81
0.94
0.55
0.91
0.68
0.98
-0.51
0.85
0.51
0.93
Numerical evaluation of the power predictions (measured 3D interpolated 3D) for Tx height 20m
The numerical evaluation of these differences is listed in the table above. According to this
evaluation again the AM algorithm has the best performance, i.e. the smallest error with
respect to the measured 3D antenna pattern. However, both the HPI and the WBI algorithm
achieve nearly the same performance.
35
Fig. 2-44: Prediction of received power using the measured 3D antenna pattern (Tx height 10 m)
Fig. 2-45: Prediction of received power using the interpolated 3D pattern (AM algorithm)
36
Fig. 2-46: Prediction of received power using the interpolated 3D pattern (BI algorithm)
Fig. 2-47: Prediction of received power using the interpolated 3D pattern (WBI algorithm)
37
Fig. 2-48: Prediction of received power using the interpolated 3D pattern (HPI algorithm)
The plots of the predicted received power for the reduced transmitter height of 10m show
the influence of the antenna pattern on the computation of the wave propagation. When
comparing the different power plots there are some differences visible concerning the main
lobe and also the side lobes.
Comparison
Measured
Pattern AM algorithm
Polarization +45
Mean value [dB]
Std. Dev. [dB]
BI algorithm
WBI algorithm
HPI algorithm
Fig. 2-49:
Polarization -45
Mean value [dB]
Std. Dev. [dB]
0.05
0.64
0.14
0.63
2.48
1.08
1.96
0.86
0.76
0.73
0.78
0.70
-0.32
0.63
0.29
0.72
Numerical evaluation of the power predictions (measured 3D interpolated 3D) for Tx height 10m
The numerical evaluation of these differences is listed in the table above. According to this
evaluation again the AM algorithm has the best performance. However, the HPI algorithm
achieves nearly the same performance.
38
2.5.6.3
The antenna 742 211 is a sector antenna with 67 horizontal and 14 vertical half power
beam width and 6 electrical downtilt. The horizontal and vertical patterns of this antenna
are given in figure 2-50 while the measured 3D pattern is visualized in figure 2-51.
The sketch of the 3D pattern is taken from a side view, i.e. the main lobe of the antenna
characteristic is oriented towards the right. The views in figure 2-52 represent the
interpolated 3D patterns for the four different interpolation algorithms. Similar as for the
measured 3D antenna pattern also these figures show the side view of the antenna with the
main lobe oriented towards the right.
Fig. 2-50: 2D antenna patterns in horizontal and vertical plane (logarithmic scale)
39
Interpolated 3D Patterns
Fig. 2-52: Interpol. 3D patterns by AM (top left), BI (top right), WBI (bottom left), HPI (bottom right)
40
Fig. 2-53: Prediction of received power using the measured 3D antenna pattern (Tx height 20 m)
Fig. 2-54: Prediction of received power using the interpolated 3D pattern (AM algorithm)
41
Fig. 2-55: Prediction of received power using the interpolated 3D pattern (BI algorithm)
Fig. 2-56: Prediction of received power using the interpolated 3D pattern (WBI algorithm)
42
Fig. 2-57: Prediction of received power using the interpolated 3D pattern (HPI algorithm)
The plots of the predicted received power for the transmitter height of 20m show the
influence of the antenna pattern on the wave propagation results. When comparing the
different power plots there are some differences visible, especially concerning the main lobe
and the side lobes.
Comparison
Measured
Pattern AM algorithm
Polarization +45
Mean value [dB]
Std. Dev. [dB]
HPI algorithm
BI algorithm
WBI algorithm
Fig. 2-58:
Polarization -45
Mean value [dB]
Std. Dev. [dB]
-0.26
1.84
-0.28
2.57
-0.62
1.52
0.50
2.41
0.71
2.10
0.91
2.04
-1.34
1.90
-1.12
2.05
Numerical evaluation of the power predictions (measured 3D interpolated 3D) for Tx height 20m
The numerical evaluation of these differences is listed in the table above. According to this
evaluation again the HPI and the AM algorithm have the best performance, i.e. the smallest
error with respect to the measured 3D antenna pattern.
43
Fig. 2-59: Prediction of received power using the measured 3D antenna pattern (Tx height 10 m)
Fig. 2-60: Prediction of received power using the interpolated 3D pattern (AM algorithm)
44
Fig. 2-61: Prediction of received power using the interpolated 3D pattern (BI algorithm)
Fig. 2-62: Prediction of received power using the interpolated 3D pattern (WBI algorithm)
45
Fig. 2-63: Prediction of received power using the interpolated 3D pattern (HPI algorithm)
The plots of the predicted received power for the reduced transmitter height of 10m show
the influence of the antenna patterns on the wave propagation results. When comparing the
different plots there are some differences visible concerning the main lobe and also the side
lobes.
Comparison
Measured
Pattern AM algorithm
HPI algorithm
BI algorithm
WBI algorithm
Polarization +45
Mean value [dB]
Std. Dev. [dB]
Polarization -45
Mean value [dB]
Std. Dev. [dB]
-0.73
1.73
-0.82
1.86
-0.98
1.78
0.08
1.56
0.16
2.44
0.36
1.57
-2.27
2.04
-2.09
1.84
Fig. 2-64: Numerical evaluation of the power predictions (measured 3D interpolated 3D) for Tx height 10m
The numerical evaluation of these differences is listed in the table above. According to this
evaluation the HPI has the best performance, i.e. the smallest error with respect to the
measured 3D antenna pattern.
46
Conclusions
In this section the four different interpolation algorithms are compared to one another by
analyzing the predicted power plots which are computed taking into account the
corresponding antenna pattern.
This comparison is performed for three typical antennas used in the wireless communications
domain (the antennas have different characteristics concerning the vertical and horizontal
half power beam width). Each of the antennas is used with two polarizations (+45/-45)
and for two different heights of the transmitting antenna (one location below rooftop level
and one above rooftop level). Therefore the achievement of generalizing results should be
ensured.
Taking into account all the different cases the simple AM algorithm and the HPI algorithm
have the best performance concerning the numerical evaluation. As the latter method
indicates the better agreement of the interpolated 3D pattern to the measured 3D pattern
when comparing the 3D views, this algorithm is recommended as default algorithm.
Additionally the HPI is the more generalizing approach due to the higher number of
considered parameters (i.e. more information of the horizontal and vertical pattern is
evaluated)..
The Bilinear Interpolation Methods (BI as well as WBI) produce good results concerning the
agreement of the 3D view (when comparing the measured 3D pattern to the interpolated 3D
pattern), however the accuracy of these methods for the power prediction does not reach
the accuracy of the HPI algorithm.
47
3.1
The actual antenna patterns (either 2x2D or real 3D) computed with MASC are saved in the
standard file formats of AMan and can be exported to many other file formats (e.g. *.msi,
*.pln, ). So the computed patterns can also be used together with other prediction models
and radio network planning tools.
The scenario incl. mast, arms, tubes, radoms, is modeled in 3D and automatically
generated based on some parameters (distances, lengths, materials,) defined by the user.
MASC generates not only all objects of the scenario automatically, it also generates the
coordinate system. So the user can focus on what is actually important (the properties of the
individual antennas mounted at the mast) and MASC organizes the other things.
In 2x2D mode the vertical as well as the horizontal pattern contain the direction with the
highest gain (at this point vertical and horizontal pattern intersect). Especially if antennas
with tilts are used this is very complex because the horizontal pattern is shifted horizontally.
The vertical plane is defined with two vectors and one point:
Vector from point with highest gain towards the center of the spheroid of the
computation (see section 3.5.4)
Vector of z axis
Center of the spheroid of computation (see section 3.5.4).
The horizontal plane is defined with two vectors and one point:
Horizontal vector (no z component)
Normal vector of vertical antenna pattern
point with highest gain
The computation can be scaled to obtain very fast first results and to get the highest
achievable accuracy (requiring longer computation time).
AWE Communications GmbH
48
3.2
3.2.1
From the theoretical point of view, ray optical algorithms can only be used in the far field of
the antenna and are only valid if the obstacles are large compared to the wavelength.
Additionally, they do not consider coupling between adjacent antennas.
If the near field must be computed, models based on the electromagnetic field theory are
required. E.g. the Method of Moments (or something similar). But these models suffer from
long computation times.
MASC uses ray optical methods and not the Method of Moments because of the computation
time. Short computation times are only achievable with ray optical models. And the error
concerning the accuracy is acceptable.
3.2.2
Actual radiation
Shielded
Region
Shielded
Region
Antenna 1
(transmiiting)
Object
(shileidng)
Antenna 1
(transmitting)
Object
(shielding)
Strahlende Punktquelle
Fig. 3-2: Comparison between single point radiation and actual radiation
The error shown in figure 3-2 is obviously big if the size of the antennas is large and the
distance between them is small. As soon as the distances between the antennas are
multiples of the antenna size, the error can be neglected.
AWE Communications GmbH
49
3.2.3
The materials of the mounting elements as well as walls, masts, etc. are assumed to be
isotropical and homogenous (i.e. not different electrical properties depending on the
direction of propagation in the medium).
Materials are characterized by their electrical properties (transmission loss, reflection loss).
AWE Communications offers an EXCEL sheet where these properties for different frequency
bands and materials are listed. Also the conversion from dielectricity, permeability and
conductance to transmission and reflection loss are given in this sheet.
The material properties of each element used for the mounting of the antenna (e.g. mast,
arm, sub-arm, tube, radom, wall,) can be defined individually to describe the scenario as
accurate as possible. Reasonable default values are suggested by the software.
MASC is based on the assumption, that the radom of a single antenna is included in the
pattern used for this antenna. Therefore the radom is not considered for this antenna. It is
only considered as a shielding and/or reflecting element for all other antennas.
3.3
Configurations
MASC supports antennas mounted at masts as well as antennas mounted in front of walls:
50
3.3.1
Antenna 1
Arm 1
Mast
Arm 2
Antenna 2
Arm 3
Antenna 3
Fig. 3-4: Example for Quasi-Omni Configuration
51
The difference between Quasi Omni, Skew and Side Mounted is the value of the azimuth
angle between antenna/sub-arm and arm. In Skew the init value should be 90 (i.e. main
radiation is tangential to mast), while in Quasi Omni as well as in Side mounted the init
value should be 0 (i.e. main radiation of antenna is orthogonal to mast).
Quasi Omni and Skew configurations allow multiple antennas at one mast while Side
Mounted is generally limited to one antenna.
In MASC the azimuth angles between subarms/antennas and arms can be defined
individually and the number of antennas can also be defined by the user. Therefore the user
has not to specify explicitly the configuration (Quasi-Omni, Skew, Side-Mounted ) to be used.
Only the mast configuration must be selected. By defining the azimuth angles and the
number of antennas, the user selects a configuration implicitly.
Figures 3-4, 3-7 and 3-7 show a typical Quasi Omni (see fig. 3-4), a typical Skew (see fig.
3-7) and a typical Side Mounted (see fig. 3-7) configuration.
Therefore all three configurations Quasi Omni, Skew and Side Mounted can be modeled
with the mast configuration of MASC.
Antenne
Antenne
Mast
Antenne
Antenne
Fig. 3-7: Example for Skew Configuration
Antenne
Mast
52
Side-Mounted configurations are depending on the shadowing of the mast and arms and on
the reflections at the mast and arms. Depending on the distance between antenna and mast,
different radiation directions can be generated (see figure 3-8).
Distance Antenna-Mast: /4
Distance Antenna-Mast: /2
53
54
3.3.2
The wall configuration an be used, if the antennas are mounted at or in front of a wall. Very
often only one antenna is mounted but in MASC the number if only limited by the RAM.
The wall configuration consists at least of an antenna, the arm to mount the antenna and a
wall (see figure 3-10).
Seitenansicht
Gebudewand
Seitenansicht
Gebudewand
Draufsicht
Draufsicht
Gebudewand
Gebudewand
The wall configuration considers shadowing of the wall as well as reflections at the wall. If
the transmission/penetration loss of the wall is very high, the radiation into the direction of
the wall is suppressed.
Very important is the consideration of the reflection at the wall, because the reflected rays
are superposed to the direct rays and depending on the phase difference between the two
rays, the superposed signal can be decreased or increased (see figure 3-11).
In this configuration very often only one antenna is considered. Of course, the software itself
can handle an arbitrary number of antennas. In case of multiple antennas, the orientation of
the wall is defined with the first antenna in the list of antennas.
The computation of one reflection is sufficient to obtain the required accuracy.
The size of the wall is assumed to be infinite, so diffractions at the wedges of the wall are
neglected.
55
Fig. 3-11:
Influence of wall on antenna pattern of a GSM 900 directional antenna,. The distance
between antenna and wall is 10 cm in both figures on the top and 150 mm in the figures in
the lower line.
.
The wall configuration considers the following parameters:
56
3.4
For each antenna used in the configurations presented in section 3.3, a far field radiation
pattern is available. This pattern is measured in an anechoic chamber and does not include
any masts, arms, walls, etc.. Only the radom of the antenna is included. If the antenna
consists of multiple dipoles, the pattern describes the radiation of the superposed signal
radiated from all dipoles. An isolated analysis of a single dipole is not possible. The radiation
is concentrated in a single point even if multiple dipoles are used to build the antenna (see
also section 3.2.2).
The pattern is provided by the manufacturer of the antenna and it is either 2x2D (horizontal
and vertical plane) or real 3D (see figures 3-12 and 3-13). With the algorithms presented in
section 2.5 it is possible to compute the 3D pattern based on the vertical and horizontal 2D
patterns. Figure 3-13 shows an example of such a 3D pattern based on the patterns shown
in figure 3-12.
Fig. 3-12:
Vertical (left) and horizontal (right) antenna pattern (Antenna Kathrein K739856)
3.5
3.5.1
As already mentioned in section 3.2.2, the antenna pattern is based on the assumption of
radiation from a single point (therefore they are only valid in larger distances (several wave
lengths) to the antennas. In the near filed different approaches are used to determine the
radiated field (FEM, MoM,). Due to this single point radiator concept, the propagation can
be approximated with spherical waves (and different gains in different directions).
If the distances to obstacles are several wave lengths (what is true for the frequencies in
cellular wireless (more than 800 MHz) and the configurations supported with MASC (see
section 3.3.1)), the local wave front is nearly planar and the well known equations to
compute the reflection, penetration and diffraction losses can be applied.
3.5.2
To compute the electric field strength radiated from an antenna, the transmitted signal
power must be known. In MASC the user has not to specify the power because the resulting
pattern is always independent of the absolute power (only relative power ratios between
different antennas are important). Therefore a reference Tx power of 1 W is applied for each
antenna.
If power splitters are used, the power is split between the antennas according to the settings
of the power splitter.
Example:
4 antennas fed with power splitter 1 : 1 : 5 : 5. The powers of the antennas are 0.2 W
(antenna 1 and 2) and 1 W (antenna 3 and 4).
The isotropical radiator (see next section) is also fed with the sum of all single antennas (in
the example above this would be 0.2 W + 0.2 W + 1 W + 1 W = 2.4 W reference power.
3.5.3
The isotropical radiator is used as reference for the computation of the resulting antenna
pattern. It radiates the same power density in all directions and is described in section 2.1.
As already mentioned, the isotropical radiator is a single point of radiation. The coordinates
of this point are determined based on the coordinates of the single antennas used in the
configuration. In a first step, the geometrical mean Mgeo of the single antennas (each of
them concentrated in a single point, see section 3.5.1) is determined. Each antenna is
weighted with its individual part of the Tx power (if power splitters are used). So in power
splitter mode, Mgeo will be closer to the high power antennas.
The total Tx power Pt0 fed to the isotropical radiator is the sum of the powers fed to the
individual antennas (see section above). With this power the received power Pr in a distance
r can be computed with
Pr
Pt0 Gt Gr 02
4 r
For the computation of the resulting antenna pattern the field strength and not the power is
relevant. Therefore the power density Sr is computed with
58
Sr
Pt0 Gt
4 r 2
With
Sr
Eeff
ZF 0
the power density Sr can be transformed to the effective electric field strength Eeff (ZF0 is
the free space impedance and its value is 120 ).
3.5.4
General concept
The geometrical mean Mgeo (see section 3.5.3) is used as center of the computation
spheroid. The radius of the spheroid is rK and can be very large to be sure that all objects of
the configuration (masts, walls, arms,) are inside the spheroid (the radius used is between
500 and 1000 m, depending on the objects of the configuration) but the actual value has
no influence on the accuracy of the results.
Quasi Omni
Antenne-vor-Wand
Berechnungskugel
Berechnungskugel
geo
geo
E tat
E iso
represents the gain in the direction towards the point on the spheroid currently examined.
Only the magnitude of the complex field strength is used as the phase is not written in the
result. After all points on the spheroid are computed, the resulting pattern is available.
AWE Communications GmbH
59
3.5.5
3.5.5.1
Rays
In the basic version only the direct ray from the antenna (single point, see section 3.2.2) to
the examined point on the spherical surface (see section 3.5.4) is considered. All
transmissions of this ray are evaluated.
A consideration of reflections and diffractions makes no sense because of the assumption of
a single point radiation source (see section 3.2.2). The reflection conditions (angle of
incident ray = angle of reflected ray) are seldom true (in contrast to a real antenna which is
not concentrated in a single point) and therefore nearly no reflected rays are found.
Figure 3-15 shows this effect. If the antenna is reduced to a single point, many diffracted
rays are neglected (especially if the distance between antenna and objects is small). A real
antenna (not reduced to a single point) would have reflections at the object.
Due to the fact that the
reflection loss reduces the
power of this ray, the error in
the superposition of the
reflected and direct rays can
be neglected (if 10 dB or
more reflection loss is
applied, the sum of both rays
in logarithmic scale is nearly
identical to the direct ray
alone).
Antenne als
Punktquelle
Empfnger
Sender
(Punktquelle)
Antenne
(real)
Empfnger
Sender
(Real)
3.5.5.2
Direct Ray
The direct ray from the single antennas to the spherical computation surface (see section
3.5.4) are very important. They have as many transmissions/penetrations (see section
3.5.5.3) as required.
The contribution of the direct rays to the total field strength are computed with the
equations given in section . If there are additionally transmissions/penetrations, the
additional loss due to these interactions (see section 3.5.5.3) is added to the loss of the
direct ray.
60
3.5.5.3
The number of transmissions/penetrations are not limited in the rays considered for the
computation. The transmission/penetration loss of an object intersecting any ray must be
added to the free space loss LFS of this ray,:
LFS 10 lg
20 lg
Pt0
Pr
10 lg Pt0 10 lg Pr
4 r
10 lg Gt 10 lg Gr
An object is defined as a planar polygon (see chapter 3.6). The material properties of the
objects can be defined individually (see section 3.6.9).
The determination of the intersection between a ray and an object is explained with a simple
example. If more objects must be considered, this analysis must be made for each object
individually.
As all objects are planar, they can be
represented by their normal vector nW (see
figure 3-16).
On the following pages the transmission o
the ray between the two points P1 and P2 will
be considered. The point Q is computed as
intersection between the line represented by
P1 and P2 and the object (described by the
normal vector nW and its distance to the
origin of the coordinate system). If nW is
parallel to the direction from P1 to P2, no
transmission/penetration is assumed.
P2
nW
r
Q
Objekt
P1
Fig. 3-16: Intersection of ray and object
If Q is between the points P1 and P2 and if Q is inside the polygon representing the object, a
transmission/penetration occurs and its transmission/penetration loss (material dependent,
see section 3.6.9) must be added to the free space loss (see above) of the ray between P1
and P2. If Q is outside the polygon or not between P1 and P2, no additional loss due to
transmission/penetration must be considered.
P2
Objekt
Objekt
P2
P1
P1
Fig. 3-17: Transmission/penetration must be considered or can be neglected
61
3.5.5.4
Reflected Rays
Due to the very small distances between the reflecting objects and the concentrated point of
radiation, the reflection is only considered correctly if the object is significantly larger than
the wavelength and larger than the distance between object and center of radiation. This is
not valid for most objects (except of the wall in the configuration antenna in front of w a
wall (section 3.3.2)). So theoretically the reflections at masts, arms and radoms should not
be computed.
But to obtain good and accurate results, the reflections are important. And therefore they
are included in the computation. Reflections at mast and wall are extremely important
because the reflection loss is not very high and the signal might have a phase shift of 180
to the direct signal in some directions - which leads to destructive interference and therefore
to an elimination of the radiation into this direction. So the reflections are important and
must be considered. Reflections at radoms can be neglected because their attenuations are
high and so the superposition to direct paths (without reflections) is nearly invisible in the
resulting antenna pattern.
The computation of the reflected ray will be
shown in the following example. The ray
emitted from P1 will be received by P2 and
the reflection at the wall is analyzed.
If more than only one object must be
considered, then this step must be repeated
for each object.
In a first step, the image of P1 relative to the
wall is determined. The distance dPW of the
point P1 from the wall is computed with the
normal vector nW of the wall.
P1
P2
Wand
Fig. 3-18: Scenario for the reflection
The image P1A is determined by multiplying the normal vector of the wall with -2 * dPW and
adding it to P1 (see also the two figures below):
P1A = P1 2 * dPW * nW
P1
P2
dPW
nW
Wand
Fig. 3-19: Geometry to compute the reflection
62
P1
P2
dPW
nW
Wand
dPW
P1A
Fig. 3-20: Image of the transmitter
If the line from P1A to P2 is intersecting the polygon representing the wall, this point of
intersection is named Q and represents the point of reflection.
P1
P2
nW
Q
Wand
P1A
Fig. 3-21: Computation of reflection point Q
The angle of reflection between the ray and the wall is determined with the scalar product
of the normal vector nW and the vector q of the reflected ray (from Q to P2).
P1
P2
nW
q
Q
Wand
P1A
Fig. 3-22: Computation of the angle of reflection
AWE Communications GmbH
63
cos nw * q
With the angle a and the material properties of the wall/object (see section 3.6.9) the
reflection loss can be determined accurately. More information related to the determination
of the reflection loss can be found in section 3.5.7.
This reflection loss is then subtracted from
the free space loss given by
LFS 10 lg
20 lg
Pt0
Pr
10 lg Pt0 10 lg Pr
4 r
P1
P2
r1
r2
10 lg Gt 10 lg Gr
Wand
r = r1 + r2
For both parts of the reflected ray (from P1 to Q and from Q to P2), the
transmissions/penetrations must be determined additionally (see previous section).
3.5.5.5
Diffracted rays
The explanations for the reflected rays with the size of the objects and their distance to the
radiator (see previous section) is even more important for the diffractions. And the
diffraction losses are high compared to the reflection losses.
So MASC does not consider any diffractions in the computation of the resulting patterns.
Diffractions would only be interesting if no reflecting or transmitting ray would reach the
point currently examined. But this can never be the case as an unlimited number of
transmissions/penetrations are supported and therefore at least the direct ray is examined
for each point examined.
3.5.6
For each element an individual transmission/penetration loss can be defined (in dB). If a ray
is penetrating the object, this loss is added to the propagation loss of the ray.
AWE Communications offers a table (MS EXCEL Sheet) with different materials and their
specific losses, depending on the thickness. So the user can enter the material and the
thickness and the sheet provides the appropriate transmission/penetration loss to be used.
Transmission/penetration of dielectric objects does not lead to a phase shift. Therefore no
additional phase offset is added to the phase of the ray at the receiver.
64
3.5.7
3.5.8
Polarization
The polarization of the electromagnetic wave has a strong influence on the resulting antenna
pattern. But to consider polarization in the resulting pattern, also the input patterns of the
single antennas must have information about the polarization. But this polarization data is
not available for most patterns of commercially available antennas.
So if the input does not include the polarization, it makes no sense to consider polarization in
the result. Therefore the result is independent of the polarization of the signals.
3.5.9
Superposition of rays
The individual rays are superposed with consideration of the phase information. This leads to
constructive and destructive interference in the results.
Depending on the path length and frequency, the phase of the signal and the complex
electric field vector are computed. The Phase i of ray i is depending on the path length r
and the wave length of the signal (c0 is the velocity of light and f the carrier frequency):
* 360
r
f * 360
c0
For each reflection in the ray, an additional phase shift of 180 is considered.
The total field strength Etot is the coherent (i.e. incl. phase) superposition f the electric field
strengths of the different paths:
E tat
E iso
65
3.6
In this subsection the mathematical modeling of the objects (masts, walls, arms,) to
describe the scenario is given. Only the objects relevant for the computation are modeled.
Small objects like skews, nuts, etc. are not modeled because their influence is quite small.
3.6.1
Mast
The mast is used for many configurations (see section 3.3.1) and is described with:
As the mast is a cylinder with circular cross section, the information related to circular
objects given in section 3.6.6 applies also for the mast.
3.6.2
Wall
In the configuration antenna in front of a wall (see section 3.3.2) the wall is very important
and must therefore be modeled accurately with the following parameters:
3.6.3
Antennas are mounted at arms. They are orthogonal to the surface (mast or wall) and
consist always of a pair of two arms. One above the other. At these arms a tube is mounted
vertically. At the tube there are at least two sub-arms to mount the antennas at the tube.
All elements of the arms and sub-arms are geometrically centered. The cross section of the
tubes is always circular (the information related to circular objects given in section 3.6.6
applies here). The tube is open at its ends (see figure 3-24).
Sometimes the antenna is mounted at the tube without sub-arms (see figure 3-25).
Therefore the user can enable or disable the sub-arms for each antenna individually in
MASC.
66
Mast
Ausleger
Antennenradom
A
C
Seitenansicht (a)
Draufsicht (b)
Fig. 3-24: Side-view (a) and top view (b) of the arms (A) mounted at the mast (M).
The antenna (R) is mounted with sub-arms (C) at the vertically oriented tube (B)
Fig. 3-25: Side-view of the arms (A) mounted at the mast (M).
The antenna (R) is mounted without sub-arms at the vertically oriented tube (B)
67
60
270
240
120
180
45
-60
C
Fig. 3-27: Top view of a 45 azimuth rotated antenna. Antenna (R) and sub-arm (C) are rotated in the
horizontal plane around the tube (B). The location of the arms (A) and tube (B) is not influenced.
60
R
Fig. 3-28: Top view on a configuration with two antennas (R),
mounted at one tube/arm. The azimuth of the antennas is 60.
AWE Communications GmbH
68
3.6.4
The radom of the antenna is a protection of the radiating elements. The influence of the
radom on the radiation of the antenna is already included in the pattern of the antenna but
in MASC the influence (mainly shielding) of the radom on other antennas mounted at the
same mast is considered.
The cross section of radoms are either circular cylinders (for omni-directional antennas) or
rectangular (for sector antennas). Of course the user can select the cross section of the
radoms individually for each antenna (see figure 3-29)
Fig. 3-29: Example for the radom of an omni antenna (a) and a sector antenna (b)
69
3.6.5
Antennas can have a tilt (uptilt or downtilt). Theoretically tilts with a maximum of 90 are
possible. Positive values describe uptilts and negative values are used for downtilts.
Practically tilts are most often negative (i.e. downtilts) and are in the region between 0 and
20.
A downtilt leads to a rotation at the lower end (bottom) of the antenna (see figure 3-30)
while an uptilt leads to a rotation at the upper end (top) of the antenna (see figure 3-31).
C
A
A
C
A
C
The sub-arms should be longer in case of a tilt but in MASC this effect is not modeled as its
influence on the accuracy is not visible in the output pattern.
70
3.6.6
As already mentioned several times, objects with circular cross sections can be approximated
with multiple planar rectangles (see figure 3-32) without a big influence on the accuracy of
the results. For the mast this approximation is only used for the determination of the
transmissions/penetrations (see section 3.5.5.3), i.e. for the determination of the shielding of
the mast. For the reflections at the mast (see section 3.5.5.4) the ideal circular cross-section
is used to obtain a high accuracy in the results.
At least 6 (better 10) planar rectangles should be used to approximate the cylinder
accurately.
n = 10
z
y
x
modellierter Zylinderquerschnitt
D
realer Zylinderquerschnitt
a
b
3.6.7
Objects like arms or tubes can also have a rectangular cross-section and therefore the user
can also model these cuboids with the following parameters:
Width
Depth
Height
Position (x-y-z Coordinates)
Each cuboid consists of bottom, ceiling and 4 walls. But bottom and ceiling of the cuboid are
not modeled as their influence on the resulting pattern can be neglected.
The radoms of sector antennas (see section 3.6.4) can be approximated with cuboids.
The cross section of sub-arms is always a quadrate (width and depth are identical).
AWE Communications GmbH
71
3.6.8
Positioning of Elements
The objects describing the scenario must be positioned relative to one another. This can be
defined via x-y-z-coordinates in the coordinate system. As this is very complicated for the
user, MASC generates the coordinates automatically after requesting some simple
parameters from user. All values, entered by the user, are relative values to the border of
the wall or mast or arm (see below).
3.6.8.1
Arms (at masts)
The arms are starting inside the mast (inner circle of polygon, see figure 3-33) to guarantee
that there are no gaps between mast and arm which would have a significant influence on
the accuracy of the pattern.
n = 10
modellierter Mastquerschnitt
realer Mastquerschnitt
Inkreis des Polygons
oberer Hauptarm
Mast
unterer Hauptarm
72
3.6.8.2
Wand
oberer Hauptarm
x'
Mitte
z'
unterer Hauptarm
x'
3.6.8.3
Tube (vertical)
The vertical tube is always fixed in front of the two arms (see sections before). Its vertical
center is on the height exactly between the two arms. This is identical to the description
given in section 3.6.3.
3.6.8.4
Sub-Arms
Sub-arms are optional. If they are used, they are always fixed at the tube (see section
before and section 3.6.3). Similar to the arms, the distance between the sub-arms is also
defined by the user. The sub-arms are vertically centered to the tube. Their azimuth
orientation is defined by the user and is similar to the azimuth of the antenna itself.
3.6.8.5
Radom
Omni as well as sector radoms (i.e. circular as well as rectangular cross section) are
mounted directly at the sub-arms (if used) or at the tube (if no sub-arms are used). The
azimuth and the tilt are defined as described in sections 3.6.4 and 3.6.5.
3.6.9
The material properties of each object can be defined individually. To have a simple interface
for fast computations the user has to defined only
A table with electrical properties of different materials can be obtained from AWE
Communications. Please ask if you are interested in this document.
73
3.7
3.7.1
The following configuration is examined (these parameters must be defined by the user):
Quasi-Omni configuration
Number of antennas: N = 3
o 1 Antenna at first arm
o 2 Antennas together at second arm
Diameter of mast: 500 mm
Height of mast: 10 m
Material of mast: Steel
Properties of arms and antennas:
o Arm1:
Height (mounted): 9 m
Material: Steel
Length of both arms: 300 mm
Distance between both arms: 500 mm
Vertical tube:
Cross section: Rectangular
Length (Height): 1000 mm
Width (cross section): 50 mm
Azimuth of arm: 180
Antenna 1-1:
Azimuth: 0 (relative to arm)
Elevation: 0
Phase offset: 0
Radom:
o Type: Cuboid
o Dimension: (W x D X H): 150 mm x 50 mm x 800 mm
o Material: Steel
o Arm 2
Height (mounted): 9 m
Material: Steel
Length of both arms: 450 mm
Distance between both arms: 500 mm
Vertical tube:
Cross section: Rectangular
Length (Height): 1000 mm
Width (cross section): 50 mm
Azimuth of arm: 0
Antenna 2-1:
Azimuth: 45 (relative to arm)
Elevation: 0
Phase offset: 0
74
Sub-arm:
o Width: 40 mm
o Distance between mounting points: 100 mm
o Length: 50 mm
Radom:
o Type: Cuboid
o Dimensions: (W x D X H): 150 mm x 50 mm x 800 mm
o Material: Steel
Antenna 2-2:
Azimuth: 315 (relative to arm)
Elevation: 0
Phase offset: 0
Sub arm:
o Width: 40 mm
o Distance between mounting points: 100 mm
o Length of sub-arms: 50 mm
Radom:
o Type: Cuboid
o Dimensions (W x D X H): 150 mm x 50 mm x 800 mm
o Material: Steel:
o
Power splitter: Antenna 1-1 - Antenna 2-1 - Antenna 2-2 :
2:1:1
75
3.7.2
0.125,
-0.125,
-0.125,
-0.250,
-0.125,
0.250,
0.000
0.000
0.000
0.000
0.000
0.000
0.000,
0.217,
-0.000,
-0.217,
-0.217,
-0.217,
0.250,
0.125,
-0.250,
-0.125,
0.125,
0.125,
0.000
0.000
0.000
0.000
0.000
0.000
0.000,
0.217,
-0.000,
-0.217,
-0.217,
-0.217,
0.250,
0.125,
-0.250,
-0.125,
0.125,
0.125,
0.217,
0.217,
0.217,
-0.000,
-0.217,
0.000,
0.125,
-0.125,
-0.125,
-0.250,
-0.125,
0.250,
8.700
8.750
8.700
8.700
0.025,
0.025,
-0.025,
0.025,
-0.215,
-0.215,
-0.215,
-0.215,
8.700
8.750
8.700
8.700
-0.025,
-0.025,
-0.025,
0.025,
-0.215,
-0.215,
-0.215,
-0.215,
8.700
8.750
8.750
8.750
-0.025,
-0.025,
-0.025,
0.025,
-0.550,
-0.550,
-0.550,
-0.550,
8.700
8.750
8.750
8.750
9.250
9.300
9.250
9.250
0.025,
0.025,
-0.025,
0.025,
-0.215,
-0.215,
-0.215,
-0.215,
9.250
9.300
9.250
9.250
-0.025,
-0.025,
-0.025,
0.025,
-0.215,
-0.215,
-0.215,
-0.215,
9.250
9.300
9.300
9.300
-0.025,
-0.025,
-0.025,
0.025,
-0.550,
-0.550,
-0.550,
-0.550,
9.250
9.300
9.300
9.300
8.500
8.500
8.500
8.500
-0.025,
-0.025,
-0.025,
0.025,
-0.600,
-0.550,
-0.550,
-0.550,
8.500
8.500
8.500
8.500
-0.025,
-0.025,
-0.025,
0.025,
-0.600,
-0.550,
-0.550,
-0.550,
9.500
9.500
9.500
9.500
0.025,
0.025,
-0.025,
0.025,
-0.600,
-0.550,
-0.600,
-0.600,
9.500
9.500
9.500
9.500
-0.075,
-0.075,
0.075,
-0.075,
-0.075,
-0.075,
-0.650,
-0.650,
-0.600,
-0.600,
-0.600,
-0.650,
8.600
9.400
8.600
8.600
8.600
8.600
0.075,
0.075,
0.075,
-0.075,
-0.075,
-0.075,
-0.650,
-0.650,
-0.600,
-0.600,
-0.600,
-0.650,
8.600
9.400
9.400
9.400
9.400
9.400
0.075,
0.075,
0.075,
-0.075,
0.075,
0.075,
-0.600,
-0.600,
-0.650,
-0.650,
-0.600,
-0.650,
8.600
9.400
9.400
9.400
9.400
9.400
8.700
8.750
8.700
8.700
-0.025,
-0.025,
0.025,
-0.025,
0.215,
0.215,
0.700,
0.700,
8.700
8.750
8.700
8.700
0.025,
0.025,
0.025,
-0.025,
0.215,
0.215,
0.700,
0.700,
8.700
8.750
8.750
8.750
0.025,
0.025,
0.025,
-0.025,
0.700,
0.700,
0.215,
0.215,
8.700
8.750
8.750
8.750
9.250
9.300
9.250
9.250
-0.025,
-0.025,
0.025,
-0.025,
0.215,
0.215,
0.700,
0.700,
9.250
9.300
9.250
9.250
0.025,
0.025,
0.025,
-0.025,
0.215,
0.215,
0.700,
0.700,
9.250
9.300
9.300
9.300
0.025,
0.025,
0.025,
-0.025,
0.700,
0.700,
0.215,
0.215,
9.250
9.300
9.300
9.300
8.500
8.500
8.500
8.500
-0.025,
-0.025,
0.025,
-0.025,
0.750,
0.700,
0.750,
0.750,
8.500
8.500
8.500
8.500
-0.025,
-0.025,
0.025,
-0.025,
0.750,
0.700,
0.750,
0.750,
9.500
9.500
9.500
9.500
0.025,
0.025,
0.025,
-0.025,
0.750,
0.700,
0.700,
0.700,
9.500
9.500
9.500
9.500
0.757,
0.757,
0.792,
0.764,
8.910
8.950
8.910
8.910
-0.039,
-0.039,
-0.039,
-0.067,
0.792,
0.792,
0.792,
0.764,
8.910
8.950
8.950
8.950
-0.067,
-0.067,
-0.004,
-0.032,
0.764,
0.764,
0.757,
0.729,
8.910
8.950
8.950
8.950
0.757,
0.757,
0.792,
0.764,
9.050
9.090
9.050
9.050
-0.039,
-0.039,
-0.039,
-0.067,
0.792,
0.792,
0.792,
0.764,
9.050
9.090
9.090
9.090
-0.067,
-0.067,
-0.004,
-0.032,
0.764,
0.764,
0.757,
0.729,
9.050
9.090
9.090
9.090
0.866,
8.600
-0.141,
0.760,
8.600
-0.106,
0.725,
8.600
part:
-0.550,
-0.550,
-0.550,
-0.550,
-0.035,
10.000
10.000
10.000
10.000
10.000
10.000
10.000
10.000
10.000
10.000
10.000
10.000
76
0.831,
0.725,
0.831,
0.831,
0.866,
9.400
8.600
8.600
8.600
9.400
-0.035,
-0.141,
-0.035,
-0.106,
-0.141,
0.866,
0.760,
0.866,
0.725,
0.760,
9.400
8.600
8.600
8.600
9.400
-0.141,
-0.141,
-0.035,
-0.106,
-0.141,
0.760,
0.760,
0.866,
0.725,
0.760,
9.400
9.400
9.400
9.400
8.600
-0.106,
-0.106,
0.000,
0.000,
-0.035,
0.725,
0.725,
0.831,
0.831,
0.866,
9.400
9.400
9.400
9.400
8.600
Arm
51
52
53
54
0.729,
0.729,
0.764,
0.792,
8.910
8.950
8.950
8.910
0.067,
0.067,
0.067,
0.039,
0.764,
0.764,
0.764,
0.792,
8.910
8.950
8.910
8.950
0.039,
0.039,
0.032,
0.004,
0.792,
0.792,
0.729,
0.757,
8.910
8.950
8.910
8.950
Arm
55
56
57
58
0.729,
0.729,
0.764,
0.792,
9.050
9.090
9.090
9.050
0.067,
0.067,
0.067,
0.039,
0.764,
0.764,
0.764,
0.792,
9.050
9.090
9.050
9.090
0.039,
0.039,
0.032,
0.004,
0.792,
0.792,
0.729,
0.757,
9.050
9.090
9.050
9.090
Arm
59
60
61
62
63
64
0.760,
0.760,
0.866,
0.760,
0.831,
0.866,
8.600
9.400
9.400
9.400
8.600
8.600
0.035,
0.035,
0.035,
0.141,
0.000,
0.035,
0.866,
0.866,
0.866,
0.760,
0.831,
0.866,
8.600
9.400
8.600
8.600
9.400
9.400
0.000,
0.000,
0.000,
0.106,
0.106,
0.141,
0.831,
0.831,
0.831,
0.725,
0.725,
0.760,
8.600
9.400
8.600
8.600
9.400
9.400
0.141,
0.141,
0.035,
0.141,
0.000,
0.035,
(The origin of the coordinate system is the center of ground plane of the mast. The center
can be defined arbitrarily important is only the relative location of the elements to one
another).
The coordinates of all 3 antennas (each concentrated in a single point) in the above
mentioned coordinate system:
Antenna 1-1:
Antenna 2-1:
Antenna 2-2:
x = 0,0 mm
x = -141,42 mm
x = 141,42 mm
y = -725,0 mm
y = 866,42 mm
y = 866,42 mm
z = 9000,0 mm
z = 9000,0 mm
z = 9000,0 mm
The coordinates of the (geometrical) point of radiation Mgeo in the above mentioned
coordinate system is:
xgeo = (2 * 0,0
+ 1* (-141,42) + 1 * 141,42) / 4
= 0,00 mm
ygeo = (2 * (-725,0)
+ 1* 866,42
+ 1 * 866,42) / 4
= 282,84 mm
zgeo = (2 * 9000,0
+ 1* 9000,0
+ 1 * 9000,0) / 4
= 9000,00 mm
77
Fig. 3-37: 3D view of the 64 objects describing the scenario listed before
Fig. 3-38: 3D view of the 64 objects describing the scenario listed before
AWE Communications GmbH
78
3.7.3
The spheroid for the computation (see section 3.5.4) has a radius of 1000 m. All points to be
computed and all objects are inside the spheroid. The center of the spheroid is the point
Mgeo (see step 2).
The transmitter power of the single antennas is according to the power splitter defined as:
- Antenna 1-1: 1,0 Watt
- Antenna 2-1: 0,5 Watt
- Antenna 2-2: 0,5 Watt
Transmitter power of isotropical radiator (used as reference to determine the gain of the
actual antenna configuration):
Pt0 = (1,0 + 0,5 + 0,5 ) Watt = 2 Watt
The antenna patterns of antenna 1-1, 2-1 and 2-2 are either already in 3D or they have to
be interpolated from 2x2D to 3D with the algorithms presented in section 2.5).
3.7.4
Only points of the spheroid (see step 3) are computed. In 3D mode all points on the
spheroid are examined while in 2x2D mode the computation is done the following way:
1. The points in the horizontal plane (including Mgeo) are computed
2. The max. gain in the horizontal plane is determined.
3. The points on the vertical plane, defined with Mgeo . the vector of the z-axis and the
vector from Mgeo to the point with max gain, are computed
4. The point with max gain in the vertical pattern is determined. If this point is not in
the horizontal plane (already computed in 1), the horizontal plane is parallel moved
from Mgeo to this point with max gain and the horizontal pattern is re-computed in
this plane.
The computation is made in the coordinate system defined in step 2.
An arbitrary point P has the coordinates (xP, yP, zP) and is on the spheroid used for the
computation. . The difference vector between P and Mgeo is defined as:
x xP xgeo
y yP ygeo
z z z
P geo
Standardized to a length of 1 m this leads to the vector
x
y
z
x y z
2
x
y
z
Point P and vector describe in the 3D antenna pattern the following azimuth und tilt :
AWE Communications GmbH
79
x
tan
y
180 tan x
y
180 tan x
x
360 tan
y
if
x 0, y 0
if
x 0, y 0
if
x 0, y 0
if
x 0, y 0
= acos (z)
3.7.5
The field strength of the isotropical radiator at the point P is used as a reference. The
isotropical radiator is located at Mgeo (see step 2) and has the coordinates (xgeo, ygeo, zgeo).
To get the field strength the following steps are required:
-
x x y y z z
2
geo
geo
geo
Siso
Pt0
4 d 2
with Pt0 describing the sum of all transmitted antenna powers (see step 3).
Computation of the effective electric field strength Eiso at the pixel P based on the
power density Siso and the impedance of the free space ZF0 = 120
Eiso Siso * Z F 0
2
80
3.7.6
For each antenna the measured field strength in the point P is determined. The computation
is similar to the computation for the isotropical radiator (see step 4-A in section before).
-
Distance d between P and radiation center of the antenna (xant, yant, zant)
d
-
S ant
Ptant
4 d 2
with Ptant representing the Tx power of the antenna (see step 3).
-
Eant FS Sant * Z F 0
2
Eant 0 Eant FS LT
Depending on the path length d and the frequency f, the phase of the signal is
determined (with wave length and velocity of light c0):
i
-
*360
d
f *360
c0
S ant
P2
r1
r2
Ptant
4 r1 r2
P1
Wand
81
Computation of the effective electric field strength Ean-FSt at point P based on power
density Sant and impedance of free space ZF0 = 120
2
Computation of the reflection loss LR of the reflected ray as explained in section 3.5.7.
Subtracting the transmission loss LT and the reflection loss LR leads to the actually
received field strength at point P (on the basis of the reflected ray):
Depending on the path length d of the reflected ray and the frequency f (and wave
length ), the phase of the reflected ray is determined with the velocity of light c0:
*360 nR *180
d
f *360 nR *180
c0
For each reflection (nR reflections are in the ray), an additional phase shift of 180 is
added.
The contributions of the direct and reflected rays are superposed considering their
phases (see section 3.5.9):
3.7.7
The antenna gain GP at point P is the quotient of the actual field strength (see step 4-B)
and the field strength of the isotropical radiator (see step 4-A) at point P:
GP
E Ant
Eiso
S Ant
Siso
82
4.1
Overview
The program is based on the document/view architecture which separates program data
from their representation. Different views are used to represent the different antenna
patterns.
A document represents the entity that the user generally opens with the instruction Open
from the menu File and stores with the instruction Save from the same menu. However, a
view is assigned to a document and works between the document and the user as an
interface. The view represents an image of the document on the display and interprets the
inputs of the user as processes for the document. The view also creates figures for printing
and for the preview.
This program is considered an MDI-application (multiple document interface). It means that
several documents can be opened and each document appears in its own child window. A
new item of the document class is generated for each new document opened.
The main window presents the window framework, a header with the title and a menu and a
symbol bar. It contains no views windows rather than it only frames the work area. The view
window of the respective documents is displayed in the child window frameworks (shown in
Figure 4-1).
Child Window
Main Frame
Menu Bar
Symbol Bar
View Window
83
4.2
4.2.1
2D Antenna Patterns
Types of Antenna Patterns
4.2.2
The 2D horizontal pattern is displayed with top view. The angle stretches from 0 to 360 and
is oriented counterclockwise.
Generally the direction of 0 is describing the direction with max. gain. But this is not
mandatory. Also other directions can be used. But as the direction of 0 is very often
visualized in radio planning tools it helps the network planner if the main energy is also
radiated into this direction.
4.2.3
Meaning
Antenna Pattern Horizontal Binary
Antenna Pattern Horizontal ASCII
In the 2D vertical pattern the angle stretches also from 0 to 360 and is oriented clockwise.
0 is to the sky and 180 is down to the bottom. 90 (and 270) represents the horizontal
plane.
Meaning
Antenna Pattern Vertical Binary
Antenna Pattern Vertical ASCII
84
4.2.4
And the distance between the circles represents the interval between the gain values of the
circles shown on the display (see figure 4-4 for some examples).
The update of the display can be accelerated if not the whole display is always updated
during the manual definition of the values (see section 4.2.8).
The limitation of the angle during the drawing of antenna patterns helps if the values are
defined manually with the mouse by clicking on the display. If enabled the modification of a
single value is always limited to the angle of the value and no further values are modified
(see section 4.2.8).
Fig. 4-3: Linear (left) and logarithmic (right) scale for the 2D antenna display
(same antenna in both figures)
AWE Communications GmbH
85
Fig. 4-4: Different settings for max and min gain in 2D pattern mode
(same antenna pattern used for upper and lower part of the figure.
Only settings for display are different as shown on the right
part of the figure)
86
4.2.5
Fig. 4-5: Horizontal and vertical antenna pattern (measured or data sheet)
To generate this pattern very fast, the figures of the patterns (as shown in figure 4-5) should
be scanned. The scanned bitmap should be saved in *.bmp format (standard bitmap).
Then the user has to generate a new
pattern via the menu File and the item
New 2D Antenna Pattern (see figure
4-6).
Via the menu Bitmap and the item Import
Bitmap the user has to select the
scanned bitmap of the measured
radiation pattern (see figure 4-7).
Now the bitmap is displayed on the
screen and the pattern can be drawn
onto the bitmap after adjusting the
coordinate system of AMan to the system
of the scanned bitmap.
(An example of such a bitmap is
distributed together with the AMan
software package)
87
Figure 4-8 shows the display after importing the bitmap. Obviously the coordinate systems
do not match.
The bitmap can be processed with different operations to adapt it to the display of AMan.
First the bitmap can be moved via the menu Bitmap => Move Bitmap or via the icon
in the toolbar. The bitmap can then be moved by clicking with the left button of the
mouse on the point of the bitmap (e.g. the center of the pattern) and then (while the button is
still pressed) moving the mouse pointer to the corresponding point of coordinate system of
AMan.
Fig. 4-9: Display after bitmap was moved (center of coordinate systems match)
AWE Communications GmbH
88
The bitmap can be scaled to enlarge it (if it was very small) or to resize it in any direction.
This is done via the menu Bitmap and the item Zoom factor for bitmap or alternatively the
icon
can be clicked in the toolbar.
After this selection the a dialog shown in figure 4-10 opens.
The user can enter the scale/zoom factor and can resize the bitmap. After entering a new
factor of 150% the display changes from figure 4-9 to figure 4-11.
Of course the bitmap can be restored to its original size with the icon
bar or via the menu Bitmap => Original size of bitmap.
89
Move the center of the pattern on the bitmap to the center of the coordinate system (as
described in step 1).
Select either linear or logarithmic representation of the diagram with the settings (see
figures 4-2 and 4-3 as well as the explanations on page 69).
Define the max and min gains of the display of the pattern as explained in figure 4-2. In
the example given in figure 4-11, the max gain of the pattern on the bitmap is 0 dB and
the scale is logarithmic. An interval of 2 dB is recommended for the distance between the
circles.
Finally the coordinate system can be rotated (if required). This is enabled by clicking with
the left mouse button on one of the diagram axes. If the key is pressed , the coordinate
system is rotated according to the mouse movements.
Rotations with 90 can be done via the tool
Now the coordinate system is defined properly and the pattern can be drawn as
described in next section.
90
4.2.6
After a new pattern is created as described in section 4.2.5 (with or without bitmap), the
user must first decide if it is a vertical or a horizontal pattern. This is asked after clicking with
the mouse button anywhere in the display.
The user can define the gains in different directions by clicking with the right mouse button
on the screen. This action inserts points with defined angle/gain relation (see figures 4-14
and 4-15).
If the mouse cursor is located on a point already defined and if the right mouse button is
clicked, a context menu appears to ask if the already defined value should be deleted (as it
is not possible to define to different gains for the same angle).
With the left mouse button points already defined can be moved (while the button is
pressed). The point currently selected is automatically changed to the current mouse
position if the limitation of the angle in the settings (see figure 4-2) is not enabled. If it is
enabled, only the selected point is modified.
A higher drawing accuracy enables an increasing of the zoom factor or the use of the zoom
window. All tools related to zoom operations can be found in the menu View or in the zoom
toolbar:
In the Edit menu a UNDO function is available. If a point is drawn wrong the user has the
option to remove it from the list via the UNDO command.
Especially if patterns are generated by the user (either drawn manually or entered
numerically), many points are not defined. To increase the accuracy (especially for 2x2D
3D interpolation, see section 4.4.4), the user can interpolate the pattern. The angle
increment after the interpolation between the pixels can be defined in a dialogue (see figure
4-21). Figure 4-22 shows an example of such an interpolation.
91
Fig. 4-14: Inserting sample points to describe the antenna pattern (display with bitmap)
Fig. 4-15: Inserting sample points to describe the antenna pattern (display without bitmap)
92
4.2.7
After a new pattern is created as described in section 4.2.5 (with or without bitmap), the
user must first decide if it is a vertical or a horizontal pattern. This is asked after clicking with
the mouse button anywhere in the display.
Especially if patterns are generated by the user (either drawn manually or entered
numerically), many points are not defined. To increase the accuracy (especially for 2x2D
3D interpolation, see section 4.4.4), the user can interpolate the pattern. The angle
increment after the interpolation between the pixels can be defined in a dialogue (see figure
4-21). Figure 4-22 shows an example of such an interpolation.
93
4.2.8
An antenna pattern loaded from a file or drawn manually can be modified by shifting,
inserting or deleting points. The pattern itself is defined with different points as shown in
figure 4-15. On these points the following operations are possible:
Move points
Points can be moved after clicking with the left mouse button on a point and while the
button is pressed, the mouse is moved to the new location of the point. After the button
is released the point is finally moved to this location.
During the mouse movement the angle can either be limited to the sector between the
two adjacent points or the point can be substituted with the neighboring one if the angle
sector of the neighboring point is reached. The current setting is defined with the Angle
limitation checkbox in the settings dialogue (see figure 4-2).
UNDO for movements is possible and is available in the menu Edit => Undo move point.
Delete points
To delete a point is must be selected with the right mouse button by clicking on it. A
context menu will appear and ask if the points should be deleted or not.
Rotate pattern
The values in the pattern can be rotated via the menu Edit => Rotate pattern.... After
selecting this menu item, the following dialogue appears:
The user can now rotate the values in the pattern. This is especially important if nonWinProp file formats are imported and their coordinate systems have a different
orientation (e.g. in the vertical pattern the 0 is in the horizontal plane and 90 are
towards the ground. Then the whole pattern can be rotated with 90 to have it in the
WinProp format). In the following figure an example is given.
AWE Communications GmbH
94
95
Interpolate pattern
Especially if patterns are generated by the user (either drawn manually or entered
numerically), many points are not defined. To increase accuracy (especially for 2x2D
3D interpolation, see section 4.4.4), the user can interpolate the pattern. The angle
increment after the interpolation between the pixels can be defined in a dialogue (see
figure 4-21).
96
4.3
4.3.1
2x2D antenna patterns contain the vertical and the horizontal pattern in the same file. This is
very often used in commercial file formats of the antenna manufacturers.
Meaning
MSI Planet File Format
PLANET File Format (identical to *.msi)
Depending on the version of AMan not always all file formats are supported. Please check
your version and the filters included in it.
4.3.2
If 2x2D antenna patterns are opened via the menu File => Open Antenna Pattern, AMan
opens two documents/windows. One window displays the 2D horizontal view and one
window displays the 2D vertical view.
All operations for the 2D view described in section 4.2 can be applied to the documents. The
antenna patterns can also be saved after the modifications.
But it must be mentioned, that each pattern (either vertical or horizontal) is saved separately
in a file in the 2D antenna file format of WinProp (see sections 4.2.2 and 4.2.3).
2x2D antenna patterns can also be loaded and
directly displayed as 3D pattern (more
information related to 3D patterns can be
found in section 4.4).
In the File menu the Convert MSI to 3D (for
*.msi and *.pln) command loads the 2x2D
pattern and compute the 2x2D => 3D
interpolation (see section 4.4.4). So the
pattern will be displayed in 3D view (see
section 4.4).
97
4.4
3D Antenna Patterns
3D antenna patterns describe the radiation of the electromagnetic wave in all directions
accurately. Very often the 3D patterns are available from the manufacturer of the antenna. If
not is can be interpolated based on the horizontal and vertical 2D patterns (see section
4.4.4).
4.4.1
Windows XP, Windows 2000, Windows NT 4.0 as well as Windows ME, Windows 98, and
Windows 95b contain special drivers for fast drawing three dimensional graphics. This
interface is called OpenGL and is a standard for 3D data representation.
OpenGL is actually itself a programming language which describes objects in the three
dimensional space with their graphic primitives (point, line, triangle). More complex bodies
can be realized from those primitives.
AMan uses OpenGL. So the computer must have installed OpenGL to be able to show the 3D
patterns. If problems with the 3D display occur, all (hardware) accelerations of the graphic
adapter should be disabled in the system control section (display adapter settings). In case
of any further problems please contact the support team of AWE Communications via
[email protected].
98
4.4.2
Loading a pattern from a file is possible via the menu File => Open Antenna Pattern.
Meaning
Antenna Pattern Binary
Antenna Pattern ASCII
2x2D patterns (see section 4.3) can be loaded directly in 3D display (if interpolated while
loading, see section 4.3.2).
4.4.3
Display of 3D Patterns
99
100
With the help of the control panel the display of the antenna pattern can be controlled:
1
The selection of the color to lighten the object is important to emphasize details.
Dark colors should be selected. Since the standard color of the surface of the objects
is red, the result on the display will not correspond exactly to the selected color (it is
a mix of red and the selected color).
The background color used is always identical to the color selected (no color mix).
While for the printing of the graph, a white background is reasonable, a more
impressive result can be obtained with dark backgrounds.
In order to draw the antenna pattern with wire frame, this check box must be
enabled.
Fig. 4-29: Same colors but wire frame enabled (left) and disabled (right)
101
In order to draw the antenna pattern with filled surfaces instead of a wire frame,
this check box must be enabled (be careful: higher computing time necessary if filled
objects are selected).
Fig. 4-30: Same colors but filled surfaces enabled (left) and disabled (right)
The program displays the type of scale selected (logarithmic or linear). See figure
4-25.
Fig. 4-31: Everything identical except of the linear (left) and the logarithmic scale (right)
All scale settings can be defined by clicking on this button. The dialog already shown
and explained in figure 4-25 appears.
Resolution of the display. The finer the value, the slower the display. Value below 1
make no sense as the finest resolution in the pattern itself is 1. To accelerate the
display the values can be set to 5. This has no influence on the data itself and its
accuracy saved in the file.
Fig. 4-32: Everything identical except of the angle increment: 1 (left) and 5 (right)
AWE Communications GmbH
102
The pattern can be displayed either in 3D mode or the vertical and horizontal planes
can be displayed. The user has to select the view with the radio button.
103
The user can select if the pattern and the coordinate system should be displayed or
not. This is especially important for the MASC module (see section 4.5).
Controlling the display with the mouse, the toolbar or the menu:
Apart from the instruction of the control panel the display can also be influenced with the
mouse, toolbar or menu.
Mouse operations
If the left button is clicked and pressed, the 3D view can be moved on the screen
according to the mouse movements.
With the right button of the mouse the display can be zoomed. For zooming press the
right button and keep it pressed while the mouse is moved. This changes the settings of
the zoom.
Pressing both mouse buttons simultaneously and moving the mouse rotates the antenna
pattern on the screen.
Toolbar operations
A zoom of the graph is also possible with the toolbar. From the left to the right: First a
fit-to-screen button. Second the possibility to define/draw the zoom window with the
mouse. And finally the two standard symbols to increase and decrease the zoom factor.
Menu operations
The zoom settings (increase, decrease, fit-to-screen) can also be selected in the menu
View.
104
4.4.4
A 3D antenna pattern can be loaded from a file and displayed (see sections before) or it can
be interpolated from a vertical and a horizontal pattern.
The interpolation algorithms and
their performance are explained in
section 2.5.
All four algorithms presented in this
section are included in AMan and
can be selected by the user (see
figure 4-36).
For such a conversion the
increments between the values to
be interpolated can be defined to
control the accuracy, the file size
and the computation time.
2x2D antenna patterns can be loaded as well as 3D patterns. But the pattern will always be
displayed in 3D view.
105
4.5
4.5.1
General Concept
MASC as described in chapter 3 is fully integrated in AMan. It has a standard API interface,
so it can also be integrated as dynamic link library into any other radio network planning
tool. It is written in C, so it can be compiled on each platform and operating system
providing a C compiler.
Project file vs. antenna pattern
The user has to define the scenario and
the antennas used in the configuration
together with all other parameters
described in chapter 3. All these settings
are saved in a project file with the
extension *.ank.
The menu File => New Multiple Antenna
Config allows the user to initiate a new
MASC project.
Config
In the menu Multiple Antenna Config the item Save Scenario (WallMan) is also available.
With this item the generated objects (masts, walls, arms, radoms,) are exported in
WallMan data format to be further processed and analyzed with WallMan. This might be
interesting when the material properties must be checked or if further checks are required.
106
4.5.2
Main Settings
In the previous section it was described how a new
project can be initiated or an existing project can be
loaded. After the project is initiated or loaded, the
general 3D view appears with two additional control
elements (marked with red in figure on the left):
Checkbox View => Objects
The drawing of the objects (wall, mast, arms,
radoms, etc.) can be enabled or disabled. If
displayed this helps to understand the
configuration and if not displayed it helps to
analyze the details of the pattern computed by
MASC.
Button Configuration
With this button the dialogue to define the
configuration is called (see figure 4-41).
107
The main settings (see figure 4-41) contain the following blocks:
Scenario
Here the user can select if a mast configuration or a configuration in front of a wall is
computed.
Basic Geometry
Here the geometrical properties of the mast or the wall (depending on the scenario, see
above) must be defined:
o
For both configurations the material properties must be defined (see section 3.6.9) here.
Arms
In this area the user can add arms to the mast or wall (depending on the scenario). Or arms
can be deleted if they are no longer required.
Of course it is also possible to edit the properties of an arm (see section 4.5.3).
The number of arms supported is only limited by the RAM of the PC.
Output (Computed Pattern)
The type of the output pattern can be defined. Either the very fast computed 2x2D pattern
or the full 3D pattern can be computed.
Additionally a filter can be enabled to smooth the resulting pattern. The order of the filter
can also be defined.
For debug purposes a logfile can be enabled. This should be used if a problem occurs on the
PC of the customer which cannot be reproduced at the PCs of AWE Communications. Then
the logfile should be generated and sent to AWE Communications to help the support
engineers to solve the problem.
The resolution of the resulting pattern can also be defined. The finer the resolution, the
longer the computation time and the more accurate the result. Resolutions below 1 are not
possible.
Computation parameters
To consider the phase correctly when superposing the signals (see section 3.5.9), the
frequency must be defined here.
Additionally the consideration of reflections for the mast configurations can be enabled. This
extends the computation time but improves the accuracy. In the wall configuration the
reflections are always determined.
AWE Communications GmbH
108
4.5.3
Settings of Arms
Geometry of Arms
A
The azimuth of the arm must be specified according to sections 3.6.8.1 and 3.6.8.2. 0 is
North, 90 is East. Orientation is clockwise.
In wall configurations the user must specify additionally the distance between the antenna
and the wedge of the wall (see figure 3-35).
109
Antennas
In the antenna section different antennas can be added to the arm/tube. Or they can be
deleted and removed from this arm.
And of course it is possible to edit the properties of the antenna (see section 4.5.4).
110
4.5.4
The properties of each antenna can be defined individually with the dialogue shown in figure
4-46. On the following pages the different parts of this dialogue are explained:
Antenna Pattern
Here the user can select the antenna pattern of the antenna as provided by the
manufacturer (measured including radom, but without masts or tubes,). MASC supports
only the 3D WinProp binary antenna format (see section 4.4). But as also mentioned in
section 4.4, many different formats (e.g. *.msi, *.pln, *.dat,) can be converted into
WinProps binary format with AMan.
Orientation
The user must specify the azimuth or the antenna relative to the arm. An angle of 0 means
the antennas main radiation is radial to the arm. An azimuth of 90 must be used for Skew
configurations (see section 3.3.1).
An example for two different azimuths can be found in figure 4-47 and figure 4-48.
Additionally the downtilt of the antenna must be defined. Positive values mean downtilt,
negative values lead to uptilt (see also figure 4-49).
Electrical properties
The phase offset of the antenna relative to other antennas must be defined here as well as
the relative power. The absolute magnitude of the relative power is not relevant. Only the
relation to the power values of the other antennas used in this configuration. If a power
value of 2 is assigned to the first antenna and a value of 10 to the second antenna, this
means that a power splitter of 1:5 is used for feeding the antennas.
Further information to the power splitter is given in section 3.5.2.
Some results for different power splitters and phase shifters are given in figure 4-50.
111
________
Fig. 4-47: Different azimuths for antennas (left: 0, right: +90)
____
Fig. 4-48: Different azimuths for antennas and their influence on the pattern
(both configurations have power splitter 1:1 and phase offset 0)
____
____
_______
Fig. 4-49: Different downtilts for antennas (left: 0, center 5, right: +10) and their pattern
(both configurations have power splitter 1:1 and phase offset 0)
112
Fig. 4-50: Different sets of power splitters and phase shifters for the same antenna configuration
and their influence on the resulting antenna pattern
Sub-Arm Properties
The user can decide if the two subarms between the tube and the antenna itself are
modeled or not. Figure 4-51 shows the modeling with and without the sub-arms.
__________
Fig. 4-51: Modeling with (left) and without (right) sub-arms
The length and the diameter of the sub-arms can be defined. And the vertical separation of
the sub-arms (i.e. the distance between the mounting points) can be defined.
Of course the material properties can be defined (as already known from the arms, see
figure 4-45).
113
Radom Properties
The radom must be specified by the user. Two different types of radoms are supported (see
figure 4-52): Cylinders with circular ground plane and cuboids with rectangular ground
plane.
Fig. 4-52: Example for the radom of an omni antenna (a) and a sector antenna (b)
The type of the radom (circular vs. rectangular) must be selected by the user. And the
height as well as the geometry of the radom (diameter, width,). For circular radoms the
number of corners to approximate the circle can be defined (the higher, the more accurate
but also the longer the computation times. See section 3.6.6).
Of course the material properties of the radom can also be defined (as already known from
the arms and sub-arms, see figure 4-45). The shielding of the radom is not considered for
this antenna as it is already included in the antenna pattern of this antenna. But its
shielding is considered for all other antennas.
114
4.5.5
Computation
If the main dialog of the MASC settings (see figure 4-41) is confirmed with OK, the
computation starts.
If the computation is done (depending on the scenario and computation mode this could
take some minutes), the result is displayed immediately.
4.5.6
Display Options
The display is similar to the display of the 3D antenna patterns (see section 4.4). Two
additional elements to control the display in MASC mode are added and explained in section
4.5.2.
If only 2x2D instead of 3D is computed, the horizontal and vertical pattern can be displayed.
The 3D pattern is not displayed. This 3D pattern view is only possible, if the 3D computation
is available.
4.5.7
As already explained in section 4.5.1, the MASC project settings are saved with the File =>
Save menu and loaded with the File => Open Multiple Antenna Config.
The MASC project (*.ank) and the
resulting antenna pattern are different
files.
In the MASC project only the
configuration is saved and described.
The resulting (computed) antenna
pattern (*.apb) is independent of the
project settings (*.ank) and must be
saved separately via the menu Multiple
Antenna Config => Save Antenna Pattern
to be used in other radio network
planning tools.
In the menu Multiple Antenna Config the item Save Scenario (WallMan) is also available.
With this item the generated objects (masts, walls, arms, radoms,) are exported in
WallMan data format to be further processed and analyzed with WallMan. This might be
interesting when the material properties must be checked or if further analysis with ray
tracing tools are required.
115
4.6
4.6.1
File Formats
ASCII File Formats for 2D Patterns (horizontal and vertical plane)
ASCII files describing either the horizontal or vertical antenna pattern can be read with any
ASCII text editor and with AMan. These files are therefore ideally suited to exchange
information about antenna patterns with any other software package.
As already described in section 4.2, two different types of files are supported:
Extension
*.aha
*.ava
Meaning
Antenna Pattern Horizontal ASCII
Antenna Pattern Vertical ASCII
*.ava
* Comments
* The number of pairs
16
* Theta Attenuation
0.00
-18.00
15.00
-20.00
30.00
-22.80
50.00
-27.00
55.00
-32.00
90.00
-37.00
120.00 -32.00
150.00 -27.00
180.00 -30.00
190.00 -40.00
245.00 -30.00
255.00 -28.00
270.00 -24.50
285.00 -22.00
300.00 -20.50
330.00 -18.00
Further information
see section 4.2.2
see section 4.2.3
*.aha
* Comments
* The number of pairs
16
* Phi
Attenuation
0.00
-18.00
15.00
-26.12
30.00
-32.72
45.00
-37.23
55.00
-39.12
90.00
-39.72
110.00 -39.12
140.00 -37.23
180.00 -38.48
210.00 -39.86
240.00 -38.61
250.00 -37.80
270.00 -35.07
285.00 -31.24
300.00 -27.63
330.00 -18.00
In *.ava and *.aha files, the first value after the comment lines, which may begin either with
* or #, is the number of the following pairs of values.
The pairs must be in one line and consist always of an angle and a gain information. No
information about phases is stored. A new pair must be in a new line.
The pairs must neither be ordered in a special way nor they must have a constant angle
increment. All undefined angles are interpolated (if required during computation).
Angle information is stored in degrees and the gain/attenuation in dB. The gain/attenuation
is relative to the isotropical radiator and positive values indicate higher gain (compared to
iso) and negative values indicate attenuation (compared to iso).
The coordinate systems can be seen in figure 4-55.
116
4.6.2
ASCII files can also be used to describe the 3D antenna pattern. These files can be read with
any ASCII text editor and with AMan. Therefore these files are ideally suited to exchange
information about antenna patterns with any other software package.
As already described in section 4.4, the following file is supported:
Extension
*.apa
Meaning
Antenna Pattern ASCII
Further information
see section 4.4
Comments are defined by the first character in a line. If a line starts either with # or *,
AMan assumes a comment line.
The triples must be in one line and consist always of two angle (horizontal and vertical) and
one gain information. No information about phases is stored. A new triple must always be in
a new line (one line can only contain one triple).
The triple must neither be ordered in a special way nor they must have a constant angle
increment. All undefined angles are interpolated (if required during computation).
Angle information is stored in degrees and the gain/attenuation in dB. The gain/attenuation
is relative to the isotropical radiator and positive values indicate higher gain (compared to
iso) and negative values indicate attenuation (compared to iso).
The coordinate systems can be seen in figure 4-55.
= 0
= 90
= 0
x = 90
Fig. 4-55: Example for ASCII antenna patterns (3D)
117
*.apa
* Comments
* Comments
*
* Theta
Phi
* Vertical Horizontal
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
40.00
40.00
40.00
40.00
40.00
40.00
40.00
40.00
40.00
70.00
70.00
..
0.00
40.00
80.00
120.00
160.00
200.00
240.00
280.00
320.00
0.00
40.00
80.00
120.00
160.00
200.00
240.00
280.00
320.00
0.00
40.00
Attenuation
relative to iso
-17.9900
-17.9900
-17.9900
-17.9900
-17.9900
-17.9900
-17.9900
-17.9900
-17.9900
-19.3000
-20.4500
-20.3400
-20.0000
-19.5600
-19.7600
-19.9600
-19.7100
-19.0300
-23.8800
-24.5100
118
119