Notes on semiempirical terrestrial wave propagation modeling for macrocellular environments - comparisons with measurements

Due to their low computational requirements, two-dimensional semiempirical wave propagation algorithms are still indispensable tools for terrestrial mobile communications and radio broadcasting coverage predictions in the very high and ultrahigh frequency bands. In this paper, we discuss various improvements of an algorithm based on digital terrain data. The algorithm combines the basic empirical propagation curves of Okumura et al. (1968) with terrain adaptive propagation curves derived from Fresnel zone clearance analyses of several terrain subsections. Also, a multiple knife-edge diffraction algorithm with approximate evaluation of the Kirchhoff diffraction integrals is applied when the line of sight is obstructed. Prediction quality is further improved by a robust algorithm for determination of effective transmitter antenna heights, and land usage along the terrain profile different from "open terrain" is considered by a multilevel Fresnel zone blockage evaluation. Prediction errors with standard deviations of about 6 to 7 dB were found for a great variety of measurements in flat and moderately undulating terrain. Transmitter stations with very high antennas (more than 200 m) caused larger standard deviations of the prediction errors. However, improvements of the predictions could be achieved in these cases by evaluating street orientations with respect to the direct path.