Model Tuning

2^nd Phase of Radio Network / RF Planning

Once Continuous Wave (CW) testing is over and we are having required samples of field data then next step of Network/ RF planning is Model Tuning.

In this blog we will discuss how we do Model Tuning in Atoll or other Planning Tool.

This blog contains basic theory of Model Tuning and finally steps involved to do the Model Tuning in Planning Tool.

Emphasize on to understand Model Tuning in a very simpler way through step by step approach of doing it in a Planning Tool like Atoll.

If you still not going through my earlier blog on CW testing you should read it here, before going to Model Tuning phase.

Basic Understanding of Model Tuning

1^st of all we should understand the basic of Model tuning, why do we do Model Tuning.

In simple language Model Tuning is done to match the coverage distance of transmitted signal, in field and in Planning Tool for the same frequency and for similar kind of terrain and clutter.

Suppose in field we have seen through our CW testing samples that in 2100MHz in a particular Urban area Received Signal Level (-93dBm) can travel a distance of 500m.

Now our target is to set this 500m distance (for signal level -93dBm) in Planning Tool for a site which operates in 2100 MHz Band in similar Urban clutter.

Here I am considering the same example as in Continuous Wave Testing (CW Testing) blog.

Once Model Tuning is completed, a propagation Model is set and all future coverage prediction will be done based on the DESIGNED MODEL for a network.

Model will be differently set for 2G, 3G, 4G-LTE or even for 5G based on use of different frequency band and Modulation techniques.

Propagation Model and Formula:

There are different types of Propagation Model based on different terrain and frequency range.

Widely used Propagation Models are Okumura-Hata Propagation Model, Cost-Hata Propagation Model, ITU-529-3 Propagation Model and so on.

Different Model has different characteristics.

The Standard Propagation Model is a Propagation Model based on Hata Formula and is suited for predictions in 150 to 3500 MHz band over long distance (from one to 20Km).

Formula for Standard Propagation Model:

Before seeing the formula just remember one thing, formula may be complicated but the actual process of doing it practically, is not that much complicated. It is simple. We should not bother much about formula. But before doing something at least we should know the theory why we are doing all these activities.

Though this formula contains lot of factors, but almost all factors are known to us.

E.g. We know Transmitted power how much we will transmit from a transmitter.

Htx is height of the transmitter i.e. the Mobile Tower (Antenna Height e.g. 30m,40m etc.)

Hrx is Height of receiver i.e. height of a human (UE) generally considered 1.5m and like this.

Main 3 factors that we don’t know is K1 (Slope), K2 (Intercept) and D (Distance) between transmitter and Receiver.

It makes a plot called Regression Curve where we take the Standard Deviation value.

X-axis for the plot is Distance and Y-axis is Received Level (Rx Level).

Regression Curve

This curve is not exactly a straight line. Our main activity is to set different values for K1, and K2 to meet this kind of approximately straight line curve.

Setting of K Values:

So whole Model tuning is done by varying the values of K1 and K2 to get a required Model.

We can do variations in other K values also for up to K7. But usually we prefer to change K1 and K2 values only.

Setting of K1 and K2 to get a final Model is totally a trial and error method.

Once in the Planning Tool, the required distance get covered (from the above mentioned example, e.g. 500m) for a particular frequency band (e.g. 2100 MHz band) in a particular clutter (e.g. Urban) for a particular received level (-93 dBm) then Model Tuning of one clutter (In this case Urban) is completed.

Same process will be repeated for other clutter also like Dense Urban or Rural. Only difference is based on clutter the travel distance of -93dBms signal will increase or decrease from 500m.

E.g. this distance will increase to 1.5Km in case of Rural instead of 500m in Urban.

Process of Doing Model Tuning:

Before start doing setting of K values we have to create different Projects in the Planning tool for different Technology.

Separate Projects required creating for 2G, 3G, 4G-LTE or for 5G.

In each projects different parameters set for different technology.

E.g. Modulation techniques will be different for GSM, UMTS, LTE and 5G.

Suppose 2 technologies operates in same frequency band, e.g. in 1800 MHz band we are operating both 2G and 4G.

In this case though both technologies are operating in the same frequency band, yet their modulation technique will be different.

So the planning tool will do the coverage prediction based on their defined project and though both are in same band still the coverage distance for same signal strength will be different.

Sample Types:

Samples through CW Testing:

In conventional CW testing collected samples shows only the Carrier Signal Strength.

These are unmodulated signals. It contains only RX Power and Latitude & Longitude of sample collected locations.

With these samples we can’t differentiate technologies. It will give same details for 1800MHz band regardless of 2G or 4G.

Samples through existing SIM:

If we collect samples through a SIM of any existing network, collected samples are modulated signals.

So we will get a difference of coverage in collected samples for same frequency band but for different technology.

In our case suppose 1800MHz band for 2G and 4G same signal level but different travel distance.

Step by step process:

• Data/ Samples collected through conventional CW testing or through Drive Test (with Existing SIM) are imported in Planning Tool like Atoll.

We can import the same in other planning tools too.

Let’s see 1 sample data as we mentioned in CW testing.

• Here we have seen that a signal level of -93 dBm is travelling to a distance of 500m in an Urban area (Say Frequency band 2100MHz for 3G).
• Now we have plot a site in Atoll Planning Tool at the same Latitude and Longitude where we set our Transmitter/Mobile Tower during CW testing.
• We already made a Project for 3G in Atoll with all different parameters. Main difference of Parameter will be Modulation Techniques for different Technology.
• So now Atoll itself has some “K” values for the created Project.
• Run a prediction for the sample site by selecting only 1 site.
• The tool will show its own coverage area for the site based on the data available in the Project.
• This prediction value may or may not match with the actual data we have collected in field.
• In Planning tool let us define 3  signal level category

Green: -72 dBm

Yellow: -72 to -93 dBm

Red: -93 to -120 dBm

In Tool we can define any no of Signal Level category based on our requirement.

For simplicity here we are ignoring the GREEN level and will consider YELLOW for -93 dBm.

And also keeping the 3 sectors of the site in 0/120/240 degree.

• For example we consider that the prediction given by planning tool shows that -93 dBm signal i.e. YELLOW travels up to 900m distance.

• Now our main task is to change the “K” values mainly K1 and K2 until we reduced the distance of -93dBm signal from 900m to 500m for this particular site.

Sample Values for K:

Some Possible values of K are-

K1 Value is depends upon Radio Frequency and Radio Technology:

• Changing of K values are not like 1 time activity.
• It is a trial and error method.
• We would vary the “K” values until we reach the value of 500m distance coverage for -93 dBm signal.
• Every time after setting new “K” values we will run a Prediction Coverage.
• Then will check what the travel distance is.
• This Process will be repeated until we find some “K” values which shows a prediction of 2100MHz band, signal level -93dBm is covering up to a maximum distance of 500m.

• These values like 500m, -93dBm will be as per our requirement. These are not some specific values.

These are taken here just for example purpose only. E.g. for dense Urban 500m will change to 300m.

So accordingly we will do all “K” value settings so that -93dBm (or -73 dBm as per our requirement) signal level would travel up to 300m only.

So by repeating the process of “K” value change we can finally reach to a DESIGN in Tool where coverage of signal level is same both in actual field and also in Planning Tool.

This is MODEL TUNING.

Once one design is done for a clutter (in this example Urban) we will do the same process for all other different clutters i.e. Dense Urban, Semi Urban and Rural.

When DESIGN is completed for all types of clutter the Model Tuning for the whole Network is completed.

Now we can put “n” no of sites in the tool defining which sites will fall in which clutter type (Urban, Rural etc.). The Tool will give an appropriate coverage for the whole Network.

So now we can plan how many sites will be required to cover a particular town or area based on the prediction plot.

Here is a sample setting of Clutter and Parameters for a Design of 2100MHz 3G in Urban area:

I am not saying that this is an ideal / correct setting for 3G- 2100 MHz band in urban area.

This is one example I did during my 3G Greenfield project to get the required design.

In a separate article we will discuss another method of Model Tuning for an existing Network with the help of MRR report.

Do you think there are some points I missed here? Leave a comment below.

Thank you for reading so far. You are important for me, please leave your footprint here:

Email address:

Leave this field empty if you're human:

<script async src="https://pagead2.googlesyndication.com/pagead/js/adsbygoogle.js?client=ca-pub-2754967439384345"
     crossorigin="anonymous"></script>