Implementation

The model was completely implemented using MATLAB. The two path-loss exponents were obtained from data in cells A, B, C, and D using the method of least squares (to minimize standard deviation). To obtain these, the diffraction ‘indices’ of cells A, B, C, and D are first computed. Once the two path-loss exponents have been obtained from A, B, C, and D, they are used to predict received signal strength of any given cell using the diffraction ‘index’ computed for that particular cell. Note that each index is averaged over the 4 cells A, B, C, and D.

 

Diffraction ‘Index’ Computation – The function findDandK computes the diffraction indices using equation 3 for 4 cell sites A, B, C, D (If power predictions of cells E, F, G, and H are desired, diffraction indices are computed for cell sites E, F, G and H) and stores them in a vector. To find the two elevation levels as mentioned earlier, Chess-Board filtering can be applied for a cell site in the following way.

-Every alternate element in the ‘terrain’ matrix is averaged out starting with the first element in the matrix. This gives one of the two elevation values.

-The same operation is performed again, except that the first element considered is the second element in the matrix. This gives the second elevation value.

NOTE  - THE TERRAIN VALUES ARE NOT BEING MODIFIED IN ANY WAY. This is just a calculation to evaluate ‘u’ in Equation 4.

 

This done, ’u’ will simply be the absolute value of the difference between the 2 elevation values. Now ’u’ can be used to find the lower limit of the integral in equation 3 by equation 4.