Microwave Radio Terrain Diffraction Modeling
 I. Introduction

II. Model Used

III. Testing / Results

IV. References

1. Testing

The model was tested using known sets of terrain and received power data.  A special function was created and used to calculate the error:

function [ map,err ] = calculateall( cellinfo )
map = processdata(cellinfo);
comparemeas(cellinfo.meas,map);
err = (double(map) - double(cellinfo.meas)) .* (cellinfo.meas ~= 0);
figure; imagesc(err); axis equal; axis xy; colorbar; title(['Error']);

This produced error maps which were easily readable and showed where the largest errors occurred.  A sample output of this function, taken from Cell Site A, is seen below.

2. Results

>> CompareMaps('Measurements_ABCD.mat','test.mat');
Comparing Measurements in MEASUREMENTS_ABCD.MAT to Predictions in TEST.MAT


Site Cell_Info_A, mean of -3.90 dB, std dev of 6.94 dB
Site Cell_Info_B, mean of +0.85 dB, std dev of 8.26 dB
Site Cell_Info_C, mean of +5.06 dB, std dev of 7.72 dB
Site Cell_Info_D, mean of -3.34 dB, std dev of 6.82 dB

Average Error of (mean = 3.3 dB, sigma = 7.4 dB) for 4 site(s)

Raw Performance Score (RPS): 7.6

While each map was close to the true values, there were errors that cannot be neglected.  When the error map from each cell site was compared to the respective elevation maps, it was noticed that the largest errors occurred near dramatic changes in elevation.