Online radio planning

User Tools

Site Tools


Using clutter

The system has 3 types of clutter which you can use to enhance accuracy, as well as LIDAR. The simplest type, random clutter, needs no demonstration as it simply raises the surrounding terrain by a user defined amount.

Here we have a coastal site next to a large city, San Francisco. Using just the SRTM terrain data, you will see accurate coverage out to sea and to the mountains but the coverage over the city will be overly optimistic due to a lack of clutter within the SRTM data. Concrete buildings stop RF so it would be fatal to not factor them in, especially if the coverage over the city is of interest.

Over optimistic prediction without clutter

Point clutter

KML Point clutter is ideal for future construction projects where data does not exist

To use the point clutter feature, create a KML overlay in Google earth (or equivalent GIS system) containing either points, lines or a polygon. Each point should have a height which is used by the system to represent a small obstacle like a building or a wind turbine. Upload the KML file (Make sure you save as .kml not .kmz) via the upload button next to the Point clutter option in the 'Environment' menu. Now within the web interface's 'Environment' menu, select 'Point clutter' to use the (private) KML clutter you uploaded to the database. You should see speckles with radio shadows in your coverage caused by your obstacles. The higher the obstacle, the deeper the shadow. In this case, obstacles have been placed in the middle of the bay.

Landcover clutter

The University of Maryland/NASA Landcover data in the system will raise the terrain according to the terrain's classification. Here the 'Landcover clutter' button is switched to ON and the prediction repeated. Now the coverage is drastically reduced over urban areas which is to be expected as the tower is not very high and the city is dense.


The most accurate data available for RF planning is LIDAR, generated by airborne laser survey. Due to the precision way it's collected by aircraft, it's resolution far exceeds space based radar products and makes it ideal for Urban RF planning. Since 2015, CloudRF's open source engine has supported high resolution LIDAR tiles in ASCII grid format. Many cities in the UK are mapped to 2m accuracy and tiles have been loaded into the system on a request basis for other countries. Testing has proved 50cm accuracy but this is not practical from a hosting or execution point of view so the standard LIDAR resolution is 2m. For users with an expert plan, you can check LIDAR coverage by placing your cursor in an area, like the Napa valley in California which exists in the system, then clicking the LIDAR button. If a tile exists (London, Sheffield, California...) then it will be overlaid on the map. The server will automatically use this tile for planning so long as you are within it. Creating a LIDAR coverage plot is relatively slow compared with SRTM data so be patient. It is recommended you start with a radius of 1km to get a feel for how long the high accuracy predictions take.
When placing an emitter on a tall building, manually enter the GPS co-ordinates of the roof and ignore the satellite imagery as it shows tall buildings at an oblique angle so you risk missing the roof by placing it visually.

Preparing a LIDAR tile

LIDAR is a technique not a format so you will find a variety of LIDAR file formats. CloudRF uses the open ASCII Grid (Also known as ESRI grid) format which is very easy to work with. To prepare a tile fit for use with the system, convert your existing 3d raster data as follows:
  • Projection needs to be WGS84 / EPSG4326
  • Scale (cellsize) for each point needs to be between 2 and 10m
  • Co-ordinates need to be in decimal degrees, WGS84
  • The header will define xllcorner, yllcorner, xurcorner and yurcorner bounds of the tile
  • NODATA_value will be 0
The GDAL suite of tools is highly recommended for this purpose.
An example header for a working LIDAR tile (Cardiff city UK) should look like this:
ncols        5000
nrows        5000
xllcorner    -3.29575481117924
yllcorner    51.4218401957316
xurcorner    -3.15423641576402
yurcorner    51.5132387803528
cellsize     2
NODATA_value  0
 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 ... 
If you need help converting LIDAR tiles and would like yours adding to the system please contact support. A fee may be requested if substantial re-formatting is required. Priority will be awarded to tiles which are correctly formatted like the example.

Warping projection

gdalwarp -s_srs epsg:27700 -t_srs epsg:4326 -co "TILED=YES" -co "TFW=YES" -co "COMPRESS=LZW" -te $WEST $SOUTH $EAST $NORTH -ts 5000 5000 myrasterdata.tiff myrasterdata.4326.tiff

Converting a Tiff to ASCII Grid

gdal_translate -of AAIGrid -ot Int32 myrasterdata.tiff myrasterdata.asc

using_clutter.txt · Last modified: 2016/09/15 21:25 by alex