Longley-Rice model precision in case of multiple diffracting obstacles

Abstract

Radio propagation models are very important in wireless communications. If obstacles exist within the Fresnel zones, knife-edge and round obstacle diffraction theory is used to predict path loss. A standard approach for predicting multiple knife-edge diffraction is to use simple geometrical constructions within the path geometry and thus to calculate an approximate total diffraction loss. In this paper, comparisons are presented between accurate field-strength measurements and simulation results derived from, the ITM coverage prediction model (Irregular Terrain Model), also known as NTIA-ITS Longley-Rice model, in conjunction with the 3-arc-second SRTM (Satellite Radar Topography Mission) terrain data. The Longley-Rice model is used up to day almost exclusively from the FCC in the US. Comparisons are extended to the single knife-edge model and the multiple knife-edge approaches developed by Epstein-Peterson, Deygout, and Giovaneli. Inaccuracies and shortcomings of the Longley-Rice model in several multiple obstacle knife-edge type propagation profiles are indicated in the VHF and UHF frequencies. A Rohde & Schwarz FSH-3 portable spectrum analyzer with tracking generator (100 kHz – 3 GHz), and a ± 0.7dB accuracy (factory calibrated), was used in our measurements. Also, two high-precision calibrated biconical antennas: Schwarzbeck SBA 9113 and BBVU 9135, with ± 1.0dB accuracy (factory calibrated), a precision log-periodic USLP9143 with approximately 6–7 dBi gain, a commercial Iskra P20 UHF-TV band log-periodic with a 6–7 dBi gain, and a low-loss, 1.8m long, cable Suhner GX-07272-D with N-type connectors were used. For simulations we used the Longley-Rice model (ITM) incorporated into the Radio Mobile program and SRTM terrain maps. We also used the Epstein-Peterson, Deygout, and Giovaneli methods. A program, which calculates diffraction losses from the above mentioned methods and uses the SRTM maps was written in Matlab. If there are no intersection points between the 0.6F zone and obstacles, it calculates Free Space Loss, if there is one obstacle it calculates diffraction loss using single knife-edge theory and if there are two obstacles its computes diffraction losses using Epstein-Peterson, Deygout, and Giovaneli methods. Fig. 1 shows double knife-edge diffraction, while Fig. 2 shows errors of various models in dB.