CW Model Tuning
CW Model Tuning
CW Model Tuning
1 General
A theoretical radio network planning should be proven by practical radio field measurements. SIEMENS can provide detailed radio measurements using special equipment and the results of these measurements can be analysed and compared directly with the prediction of the planning tool computer run. If observed discrepancies exceed the specified tolerance, parameter adjustment can be carried out. The field strength measurements should be performed in typical areas. Also test site selection depends on radio frequency planning and the computer aided interference analysis to clarify frequency allocation problems. To carry out the radio field measurements SIEMENS can provide the following steps: preparing the measurement conditions based on a problem analysis carrying out the drive tests evaluation of measuring data comparison of measuring data with the result of calculated radio prediction tuning of the propagation model analyse of radio coverage situation For more detailed information please refer to the following chapters.
2 Scope of work
2.1 Preparing the measurement conditions based on problem analysis
It is of prime necessity fixing the data of transmitter location (geographic position, height above sea level, height above ground, etc.) as well as conditions of the test scenario (transmitting power, antenna type, chart orientation, frequency). Secondly the measured area must be specified. It is helpful to clarify the drive routes before starting the measurements. Before starting the measurement campaign the invested area shall be clarified. Especially for model tuning it is helpful to restrict the measurement routes to uniform clutter areas. This additional effort preparing the measurements spares a lot of time in post processing if the clutters are not set up exactly in the TORNADO database.
Industrial
Urban
2.2
The transmitter location shall have the same requirements like a typical site for base station, e.g. height, surroundings. The transmitter system consisting of signal generator, power amplifier, cabling and antenna is mounted on top of the roof. Generally an omnidirectional antenna is used with 5-8dBi. The antenna shall be placed on a variable mast to prevent that any obstacles limit the propagation in the near field. To reduce interference and increase the receiver sensitivity the signal bandwidth shall be small (around 10kHz). Normally a frequency-modulated signal is generated with 1kHz deviation and 2kHz shift.
omnidirctional antenna
signal generator
Nevertheless it is also possible to measure existing sites, but the results can be faulty cause of the following points: Mostly the EIRP isnt exactly defined. There is the possibility of small differences in planning and installation (e.g. shift of antenna orientation or downtilt). It is recommended to measure only in the main beam.
2.3
At first the transmitter must be installed and all transmission parameters must be adjusted (power, frequency, bandwidth,...). Normally a frequency-modulated signal with a bandwidth of 6kHz is transmitted to use a lower sensitivity at the receiver system (receiving bandwidth 10kHz). The receiver system shall provide the reception of signal levels in a sufficient range (e.g. -35..-125dBm) and be able to define the bandwidth for measurements. It is to be clarified that the test receiver is adjusted to the transmitter characteristics. After finishing the measurements the interim transmitter is dismantled.
2.4
The measurement data shall be checked for failure and whether the GPS information of the driven route is continuously given. Otherwise the routes must be digitised to fit the measured data to the cartographic positions. Plots can be delivered regarding the receive level. To verify the data to prediction results in TORNADO protocol files are prepared for use in the planning tool.
2.5
Firstly the measurements must be transferred to the workstation on that TORNADO is running. In TORNADO the measurement data can be verified according its geographical position. Measurements are comparable to prediction of the test scenario.
2.6
If the prediction deviates to the carried out measurements the propagation parameters can be tuned decreasing the difference between both. Therefore error curves can be generated presenting the deviation between measurement pixels and the prediction. For this purpose a detailed analyse can be also delivered giving average mean error and the standard deviation.
2.7
By means of the tuned model the existing network state can be predicted and documented by plots of different kind. These steps may help for a first optimisation of the radio network part.
GPS
BS / Test transmitter
Measurement car
Evaluation tool
Evaluation + Interface to TORNADO
TORNADO