Planet General Model Technical Notes
Planet General Model Technical Notes
Notice
This document contains confidential and proprietary information of Mentum S.A. and may not be copied, transmitted, stored in a retrieval system, or reproduced in any format or media, in whole or in part, without the prior written consent of Mentum S.A. Information contained in this document supersedes that found in any previous manuals, guides, specifications data sheets, or other information that may have been provided or made available to the user. This document is provided for informational purposes only, and Mentum S.A. does not warrant or guarantee the accuracy, adequacy, quality, validity, completeness or suitability for any purpose the information contained in this document. Mentum S.A. may update, improve, and enhance this document and the products to which it relates at any time without prior notice to the user. MENTUM S.A. MAKES NO WARRANTIES, EXPRESSED OR IMPLIED, INCLUDING, WITHOUT LIMITATION, THOSE OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE, WITH RESPECT TO THIS DOCUMENT OR THE INFORMATION CONTAINED HEREIN.
Trademark Acknowledgement Mentum, Mentum Planet and Mentum Ellipse are registered trademarks owned by Mentum S.A. MapInfo Professional is a registered trademark of PB MapInfo Corporation. RF-vu is a trademark owned by iBwave. This document may contain other trademarks, trade names, or service marks of other organizations, each of which is the property of its respective owner. Last updated June 10, 2010
Introduction
The Planet General Model is a good propagation model to use for macro-cell planning. It is best used for frequencies between 150 and 2000 MHz where the distance between the transmitter and the receiver ranges between 1 and 100 kilometers. Ideally, when using this model, the base station antenna heights should range between 30 and 1000 meters and the mobile station antenna heights should be between 1 and 10 meters. How the Planet General Model was originally implemented in Planet DMS and how this model has been implemented in Mentum Planet differ. On one hand, when Planet DMS performs predictions, it calculates the path loss for each pixel or element within the prediction area by calculating a terrain profile from the base site to that element. The profile is used by the propagation model to calculate the path loss to that point. Predictions do not include losses or gains due to antenna masks. This allows real time masking of antennas each time the prediction is loaded. The height profiles have been compensated for the effect of the Earths curvature. A radius of 4/3rds of the Earths true radius (4/3 x 6300km = 8400 km) is often used, although this can be changed in the Planet DMS Model Editor. On the other hand, when Mentum Planet performs predictions, it calculates a prediction for each pixel along radials in the prediction area. Then, using interpolation, Mentum Planet generates predictions in the prediction area on a per pixel basis. This results in better control of the calculation time/accuracy ratio; however, for this reason, there may be slight differences between the results generated by Planet DMS and those generated by Mentum Planet. To minimize these differences, you can increase the number of radials used in the prediction.
Technical Note
With the two piece model, both K1 and K2 can be assigned two sets of values. One set is used for d< distance and the other for d> distance, where distance is the distance in meters away from the base site specified in the Model Editor.
K 3 is the multiplying factor for log( H eff ). It compensates for gain due to antenna height. K4 K5 K6
is the multiplying factor for diffraction calculation. is the Okumura-Hata type of multiplying factor for log ( H eff ) log ( d ) . is correction factor for the mobile effective antenna height gain ( K 6 H eff ). is the effective height of base site antenna from ground.
d is the distance, in meters, of the receiver from the base site. Diffraction is the value calculated for loss due to diffraction over an obstructed path. The value produced is a negative number so a positive multiplication factor, K 4 is required.
K CLUTTER is the gain in dB for the clutter type at the mobile position in Planet DMS. In Mentum Planet, K CLUTTER represents a loss. H meff H eff
Base height Spot height Average height Slope Ground Reflection Slope Profile Absolute spot height
Base height
Effective antenna height ( H eff ) is set equal to the base site height above ground.
Spot height
If H 0b > H 0m then H eff = H b + H 0b H 0m
is the antenna height above ground at the base site. is the terrain height above sea level at the base site. is the terrain height above sea level at the mobile site.
Average height
The average height is defined as the height of the base site antenna above the average terrain height, calculated over the total area of the prediction. The effective antenna height ( H eff ) is set equal to average height. In Mentum Planet, the average height is a user-defined value.
Slope
The effective height of the antenna is calculated using the slope of the terrain over a specified distance up to the antenna. Figure 1 on page 6 displays the slope algorithm.
Technical Note
H eff = ( h 1 h 2 ) + ( K d )
Where h1 h2
is the ground height at transmitter + antenna height. is the ground height at receiver + mobile height.
d is the distance, in meters, of receiver from base site. K is the slope. This is calculated over a user specified distance ds from the mobile towards the base station using the difference in height over that range.
ds
Tx Heff h1 Rx h2
Slope K
The effective height of the antenna is calculated using the slope of the terrain at the ground reflection point closest to the receiver. The calculation automatically imposes a limit of 0.8 to 4 times the height of the base station antenna. The values specified for the Minimum Height and Maximum Height have no effect on the calculation if they are not within these limits. If the line of sight between the transmitting and receiving antennas is obstructed, the height of the base station antenna above ground is used.
Profile
The profile algorithm calculates an average height along the profile between the transmitter and receiver. Heff can be calculated in three ways:
using CCIR recommendations using the Okumura calculations using user-defined start and end points for the profile
There are three conditions for the distance between the point under consideration and the antenna:
The effective antenna height is the height of the antenna above the ground. An antenna mounted 30 m up on a mast at a ground height of 20 m would confer a Heff of 50 m on any pixel within 3 km along any profile. (ii) Distance to the antenna is between 3 km and 15 km
H eff = H transmitter + H ground averageheight
(iii) If distance to antenna is greater than 15 km, the equation for effective antenna height is identical to that in (ii) above. However, average height is now given by:
sum of pixel hights along profile from 3km to 15km ---------------------------------------------------------------------------------------------------------------------------number of pixels along this profile
Effective antenna height is given by the same equation as CCIR (ii) above. Again, the expression for the average height varies with the distance as follows. (i) The distance to the antenna is between 3 km and 15 km. average height is given by
sum of pixel heights along profile from base of antenna to po -------------------------------------------------------------------------------------------------------------------------------------------------number of pixels along this point
(ii) The distance to the antenna is greater than 15 km. For all points over 15 km, the average height between 3 km and 15 km is used. average height is therefore
sum of pixel heights along profile from 3 km to 15 km ---------------------------------------------------------------------------------------------------------------------------------number of pixels along this profile
You can define the start and end points of the profile, in kilometers from the antenna base.
H meff = ( h 0m + hm ) h 0b
The following figure shows how these heights are calculated.
base mobile h0b hm h0m
2d v i = c i -------------------------di ( d di )
Technical Note
Where is the wavelength and d is the terrain slope distance. A value of v i less than -0.8 indicates sufficient clearance for the Fresnel zone is obtained over the whole path. The path loss equation for line of sight is used. Where a loss is indicated, the Fresnel integral is used.
2 1 +j E----- = ---------- e ( j ( 2 ) )v dv 2 E0
vi
This is an integral and stored as a lookup table for values of - 0.8 vi < 2.0 and the loss is calculated from.
For edge B, the variables cb , d am and d ab are similarly used, as shown in the following figure.
Figure 4 Obstruction Loss, Edge B
10
Edge A Edge B cb
The following table is the look-up table used in calculating the path losses, in dB. For intermediate values, the loss is linearly interpolated.
Table 1.1 Path loss vi -0.8 -0.7 -0.6 -0.5 -0.4 -0.3 -0.2 -0.1 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 Ploss 0.0 -0.46 -1.13 -1.86 -2.64 -3.45 -4.29 -5.15 -6.02 -6.90 -7.74 -8.59 -9.42 -10.23 -11.03 -11.77 -12.50 -13.15
11
Technical Note Table 1.1 Path loss (continued) vi 1.0 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2.0 Ploss -13.85 -14.52 -15.09 -15.70 -16.25 -16.77 -17.27 -17.79 -18.20 -18.63 -18.94
Troposcatter model
The troposcatter model is generally used in the Planet DMS Microwave tool. It is set when the Use the Troposcatter Model check box in the Planet DMS Model Editor is selected and the distance between the transmitter and the point at which loss is calculated is greater than the transition distance, dt . Where dt = dh, when dh > 90.3953 Otherwise, dt = dhata
Where
dh =
and
dh a0
h pcs and h mw PCS and MW Receiver antenna heights above average terrain.This applies if the height is greater than 5m, otherwise it is set at 5.
12
L50 is the hourly median transmission loss 50% of the time (dB). f is the frequency (MHz). d is the path length (km). (d-dh)/8.5 (milliradians) - dh is defined above. and
N ( H, h ) = 20 log ( 5 + H ) + 4.343h
Where
H equals d/4000. h equals 10-62a0/8 km. a0 is the effective earth radius in km. M is the meteorological structure parameter; this value depends on the climate type which you select in the Model Editor. The values for each climate type are given in the table below. is the atmospheric structure parameters; this value depends on the climate type which you select in the Model Editor. The values for each climate type are given in the following table.
Climate M (dB) 1 39.60 0.33 2 29.73 0.27 3 19.30 0.32 4 38.50 0.27 6 29.73 0.27 7a 33.20 0.27 7b 26.00 0.27
(km-1)
13
Technical Note
L q = L 50 + c q L 90
Where
0.137h
99.9 2.41
99.99 2.90
The calculated loss is compared with the Free Space Loss along the path; if the free space loss is greater, this is used rather than the troposcatter loss.
Microwave application
When the troposcatter model is used in a microwave application, for a confidence level q above 50%, the troposcatter loss is calculated as follows:
L q = L 50 + L c c q L 90
Where
L c equals 0.07 exp [ 0.055 ( G T + G R ) ] G T is the antenna gain of transmitter in dBi. G R is the antenna gain of the receiver in dBi.
14
Clutter effects
Clutter losses/gains
The loss/gain (referred to from now on as a loss for simplicity) due to clutter is calculated as follows: Receiver
L
Base Station
Clutter losses are considered over a distance L. L is in meters and is definable. For x=0 to n Clutter Loss = K*Fn(Kclutterx)
Where
x=0 is the pixel at the mobile. x=n is the pixel that is L meters away. K is a scaling coefficient (usually set to 1.0). Kclutterx is the clutter loss from the clutter at point x. Fn() is the function for weighting the clutter losses. Currently the functions supplied are:
With the rectangular function, each clutter loss has the same weighting. With the others, clutter loss at the receiver has the highest effect. Clutter loss at n has no effect. The triangular function gives a linear decay. Exponential decays quickest near the mobile and logarithmic decays furthest from the mobile.
15
Technical Note
Clutter heights
Clutter heights can be added to the terrain height during predictions to calculate the obstructions loss more accurately. The clutter height is not added to the terrain height at the transmitter. Clutter heights are never added at the base station. The clutter separation factor is used to separate the mobile from the surrounding clutter; that is, to prevent the mobile being swamped by the clutter as a result of high diffraction losses. This is achieved as follows: Let the clutter separation be b, the mobile be at point Rx and the point on the profile b meters from Rx be Rb:
Mentum Planet will find the highest clutter height along the profile between Rx and Rb. Let this be hmax. Mentum Planet will not add clutter heights to any points between Rx and Rb. The clutter height added at Rb will be hmax. For the remainder of the profile, clutter heights will be added to terrain heights normally.
So if a transmitter is on top of a building, the antenna height must be set to the true height of the antenna plus the building height. If a clutter category is to be assigned a height then it must also be assigned a mobile-to-clutter edge separation distance as well:
16
Physical
hmax
Rx
Rb
Modelled
hmax
b Rx
Rb
This distance is used to adjust local clutter heights for use in the diffraction calculations. If this value is left at 0.0 the resultant very high diffraction causes wild losses.
17
Technical Note
hte is the effective base station antenna height. Calculate this using the Okumura recommendations. A,B,C are the coefficients dependent on d, see below. The table below shows coefficients for the effective height of base station antenna correction factor at several distances.
d (km) 1 3 5 10 20 40 70 A 0.5131 0.2433 0.3690 0.5457 2.568 4.289 4.225 B 11.68 14.42 15.60 17.75 11.89 7.019 4.830 C -23.32 -27.31 -29.94 -34.66 -30.61 -27.66 -23.23
h is the difference in 10% and 90% heights over a distance X along the profile from the receiver to the transmitter.
Where
18
10%
h 90%
Where
h is the difference in 10% and 90% heights over a distance X along the profile from the receiver to the transmitter.
Then, at a position of undulation (peak or valley): Khf (position of undulation/m) = Khf/(h/2) where h > 10m Khf (position of undulation/m) = 0.0 where h <= 10m
Then
Where
19
Technical Note
h is the height at position of undulation. The value of Khf (position of undulation) is the value that should be applied to the propagation equation.
The equation of the line is given as ha = ad +b (this line is obtained using a least squares fit): m = arctan(a) x 17.4532 (in mrad) Then, if 3 |m| 20 mrad: Ksp = Am2 + Bm + C A, B, C are dependent on d: The table below shows coefficients for inclination correction factor at several distances:
d (km) >60 =30 <10 A -0.009411 -0.013400 -0.002394 B 0.7620 0.6313 0.2057 C 0.22 -0.63 0.12
20
If there is water at the side of the base station: Kse = 0.0004542 +0.1143 +0.27for d 60km Kse = Kse30 + Coe(d-30)for 30km<d<60km Coe =(Kse60-Kse30)/(60-30) where = dsr/d as a percentage.
Base site
Mobile
dsr
Ksub is the correction value (dB). fc is the frequency in MHz. If Lp is the standard equation for the loss in an urban area in dB then for the suburban area: Lps = Lp - Ksub
Where
21
Technical Note
Kopen is the correction value (dB). fc is the frequency in MHz. If Lp is the standard equation for the loss in an urban area in dB then for the open area: Lpo = Lp - Kopen
Where
Kim is the correction value (dB). h is the height of knife edge. d2 is the distance from knife edge to mobile (km). d1is the distance from base station to knife edge (km).
d2
22
Tx
Rx
23
Technical Note
Technical overview
AMT optimizes the clutter absorption loss and K1 to K5 factors. For more information on the path loss equation for Planet General model, see Standard propagation model on page 4. To determine the K factors that can be automatically tuned, AMT performs correlation and cross-correlation tests between the predicted path loss and the log dm , D loss , and log H eff model components. The correlation factor calculations determine the model components that are similar with the actual path loss. A high correlation value (1) between a model component and path loss implies high similarity, indicating that the component can model path loss well. For example, if the correlation factor between path loss and diffraction is small (close to 0), using diffraction loss will not improve the root mean square (RMS) error of the model significantly. If you optimize the diffraction loss factor (K4), the RMS error will not be reduced by a significant amount and the optimized value for K4 might be invalid (less than 0).
24
Table 1.3 Cross-correlation tests Cross-correlation Tests Measurement Data d loss log ( H eff ) Model Component log ( d ) log ( d ) log ( h eff ) Correlation Factor p24 p35 Correlation Threshold p24T (AMT default is 0.9) p35T (AMT default is 0.9)
25
Technical Note
K1
Options Optimize Hata urban Default value Calculated by optimization For 150 to 1500 MHz: ( 44.9 3 ) [ 69.55 + 26.16 log ( f ) ] For 1500 to 2000 MHz: ( 44.9 3 ) [ 46.33 + 33.91 log ( f ) ] Hata suburban For 150 to 2000 MHz: K1 Hata rural HataUrban + 2 [ log ( f 28 ) ] + 5.4
2
For 150 to 2000 MHz: K1 HataUrban + 4.78 [ log ( f ) ] 18.33 log ( f ) + 40.94
2
K2
Options Optimize Hata value Free space User defined Default value Calculated by optimization -44.9 -20.0 Value set by user
K3
Options Optimize Hata value Free space User defined Default value Calculated by optimization -5.83 0 Value set by user
K4
Options Optimize Default value Calculated by optimization
26
Planet General Model Technical Note Options User defined Hata value Free space Default value Value set by user 0 0
K5
Options Optimize Hata value Free space User defined Default value Calculated by optimization 6.55 0 Value set by user
Clutter Offsets
Options Optimize Zero User defined Default value Calculated by optimization 0 Value set by user
27
Technical Note
From the Intercept, K1 (near) list, choose User Defined, and type -120 in the box. From the Slope, K2 (near) list, choose Hata Value. From the Effective Antenna Height Gain, K3 list, choose Free Space. From the Diffraction Factor, K4 list, choose User Defined, and type 1 in the box. From the Log(Heff) * Log(d) Factor, K5 list, choose Free Space. From the Mobile Antenna Height Factor, K6 list, choose Free Space.
8 9
In the Knife Edge section, in the Merging Distance box, type 100. Click the Path Clutter tab and clear the Enable Path Clutter check box.
10 Click the Troposcatter Effect tab and clear the Enable Troposcatter Model check box. 11 Click the Okumura tab and clear all of the check boxes.
28
12 Click the Effective Antenna Height tab, and from the Type list, choose Spot Height. 13 Edit any of the other settings as required. 14 Click OK. The propagation model is saved in the Models folder of your project.
29
Technical Note
From the K1, K2, and K4 lists, choose Optimize. From the K3 and K5 lists, choose Hata value, and set Clutter Offsets to 0. Click OK to close the dialog box, and then click OK to begin the model tuning process. Examine the mean error for each site. Note the sites with large mean errors and RMS errors (assuming that there is a minimum of 1000 points for each site). If not enough points are available for the site, the mean error estimates will be inaccurate.
If K4 cannot be optimized, or if the optimized value of K4 is less than 0.2, do the following in the Planet Automatic Model Tuner dialog box:
Type 0.5 in the K4 box Choose Optimize for K1 and K2 again Click OK to close the dialog box, and then click OK to begin the model tuning process. Examine the mean error for each site.
If you want to further tune the model that you tuned in Step 7, do the following in the Planet Automatic Model Tuner dialog box:
From the K1, K2, and K4 lists, type values that you obtained in Step 7. From the K3 and K5 lists, choose Hata value. From the Clutter Offsets list, choose Optimize. Click OK to close the dialog box, and then click OK to begin the model tuning process. Examine the mean error for each site. Clutter offsets with values less than 90 to 95% are considered to be unreliable estimates. Unless you think that these values are unreasonable, unreliable clutter values should not be used; instead, you should set their values to 0.
30
(Optional) If you want to further optimize the Planet General Model, repeat the steps described in Step 8, and do the following in the Planet General Model Parameters dialog box
Click the Path Clutter tab. On the Path Clutter tab, enable the Enable Path Clutter check box. Typical path clutter distances are 500 m to 1000 m. Click OK to close the dialog box, and then click OK to begin the model tuning process. See if there are any improvements not only in the RMS error but with the predictions.
10 Identify the sites that have a significant absolute mean error (greater than 5 dB) and see if these sites can be classified in a different environment. Develop a new model for sites that have large mean errors.
or
L p = L Hata + Clutter Absorption Loss Where Lp hb f
is the path loss (in dB) is the height of the base station above ground level (in meters)
is the distance between the base station antenna and the mobile receiver with a height of 1.5 meters
dkm
The Clutter Absorption Loss is the additional loss in dB with respect to the Hata Urban path loss. The valid range of the parameters is:
31
Technical Note
k ( f ) = 69.55 + 26.16 log 10 ( f )for f = 150 MHz to 1500 MHz k ( f ) = 46.3 + 33.9 ( 26.16 ) log 10 ( f )for f = 1500 MHz to 2000 MHz Where hb
dkm hm
You can optimize the propagation model by setting factors K1 to K5 to the Hata values (note that K4 is equal to 0 for the Hata model) and obtain the Clutter Absorption Loss using AMT.
32
Slope algorithm, it is recommended that you set K5 to 0. It is also recommended that Okumura correction factors are not used with the Slope algorithm. These recommendations are based on analyses of real measurement data.
Effects of each model component
You can observe the effect of each component in the model and check if an acceptable RMS error can be obtained from a simpler model using the following steps: 1 In the Planet Automatic Model Tuner dialog box, do the following:
Choose Optimize from the K1 and K2 lists. Choose Hata from the K3 and K5 lists. Type 0.5 in the K4 box. Type 0 in the Clutter Offsets box. Click OK to close the dialog box, and then click OK to begin the tuning process. See if there is an improvement in the RMS error.
If there was an improvement in the RMS error for K1 and K2, do the following in the Planet Automatic Model Tuner dialog box:
Type the values that you obtained for K1 and K2 in Step 1 in the respective boxes. Choose Optimize from the K4 list. Choose Hata from the K3 and K5 lists. Type 0 in the Clutter Offsets box. Click OK to close the dialog box, and then click OK to begin the tuning process. See if there is an improvement in the RMS error.
33
Technical Note
If there was an improvement in the RMS error, do the following in the Planet Automatic Model Tuner dialog box:
Type the values that you obtained for K1, K2, and K4 in Step 2 in the respective boxes. Choose Optimize from the K3 list. Choose Hata from the K5 list. Type 0 in the Clutter Offsets box. Click OK to close the dialog box, and then click OK to begin the tuning process. See if there is an improvement in the RMS error.
If there was an improvement in the RMS error, do the following in the Planet Automatic Model Tuner dialog box:
Type the values that you obtained for K1, K2, K4, and K3 in Step 3 in the respective boxes. Choose Optimize from the K5 list. Type 0 in the Clutter Offsets box. Click OK to close the dialog box, and then click OK to begin the tuning process. See if there is an improvement in the RMS error. Click the Effective Antenna Height tab. On the Effective Antenna Height tab, choose an algorithm from the Type list. Click OK to close the dialog box, and then click OK to begin the tuning process. Repeat these steps to see which algorithm displays a higher correlation and produces a lower RMS error or lower maximum error.
34
If there was an improvement in the RMS error, do the following in the Planet General Model Automatic Model Tuner dialog box:
Type the values that you obtained for K1, K2, K3, K4, and K5 in Step 4 in the respective boxes. Choose Optimize from the Clutter Offsets list. Click OK to close the dialog box, and then click OK to begin the tuning process.
35
Technical Note
36