The effect of 3D building reconstruction errors on propagation prediction using geospatial data in cyberspace

When the 3D building structures visualized in Google Earth are reconstructed using photogrametric method, errors or inaccuracies will occur to the building vertices and the building heights. In this paper the statistics of these errors are discussed and the effect of these errors on the propagation prediction results is examined in detail. It is found that our reconstruction method introduces distance errors to the building vertices and height errors to the building heights. These errors are less than 0.5 meters in 95% of the cases. The vertex error will cause an average mean error of -0.2 dB and an average standard deviation of 5.1 dB to the predicted path gains compared to the reference case. And the height error, in the cases investigated in this paper, is very small and can be ignored. These results match the observations in the literature for different propagation environments. The 3D reconstruction method is then shown to be of satisfactory accuracy in terms of propagation prediction.