Millimeter-Wave Radio Link Analysis for 5G FWA by Combining Measurements and Geospatial Data

Abstract

Fifth-generation (5G) fixed wireless access (FWA) at millimeter-wave (mmWave) can be a solution to fulfill the increasing demand of high-speed home broadband service. However, mmWave links have high propagation loss and are susceptible to blockage. In order to estimate mmWave coverage, appropriate radio link models are required. In this paper, we use measurements and geospatial data to analyze how mmWave can be utilized in Finnish rural areas. The measurements were conducted during summertime at 26 GHz. The measurements involved a crane and a mobile measurement van. The crane was used for measuring at three different transmitter heights, 30 m, 50 m and 70 m from the ground. During the summer, Finnish forests are covered with heavy foliage. We use the measurements and geospatial map to derive radio propagation models that are suitable for FWA in Finnish rural areas. The empirically-found vegetation attenuation models, Aalto1 and Aalto2, are compared to other models from the literature. Most of the other models, FITU-R, Weissberger, COST235 and KAIST1, overestimate the vegetation loss. KAIST2 is a good fit with 7.75 dB root mean square error (RMSE) value. The proposed Aalto1 model shows the best fit, with 5.98 dB RMSE. The model fit can further be improved by tuning it for a particular antenna height. Our analysis shows that, despite high vegetation blockage in the range of 40 m to 700 m, the excess loss remains below 40 dB.