Using Light Detection and Ranging (LiDAR) Imagery to Model Radio Wave Propagation

The purpose of this study was to determine if light detection and ranging (LiDAR) imagery could provide a significantly more accurate data set for modeling near line-of-sight (LOS) propagation at higher frequencies, specifically 27.810 GHz. than a USGS digital elevation model (DEM). In addition, the study tested for significant differences in LiDAR elevation data created at various resolutions ranging from 1 to 100 meters. Finally, this study examined the effects of various classification thresholds for transforming continuous signal strength measurements into LOS or non-LOS (NLOS) classifications used in determining prediction accuracy. The capability to transmit information via higher frequency wireless equipment requires a near LOS path between the transmitter and the antenna receiving the signal. USGS DEMs, commonly used in GIS programs to predict communication viewsheds (commsheds), represent the bare earth topography and do not reflect surface features such as vegetation and buildings. In actuality these surface features can significantly influence near LOS paths and therefore a data set that contains these features can greatly improve the ability to predict commshed areas. LiDAR is a form of active imagery that records both the bare-earth as well as these surface features, at a high resolution, making it well suited for wireless modeling applications. Results indicate that signal strength threshold classification has a direct influence on the accuracy of predicted commsheds across all resolutions. Secondly, LiDAR resolutions lower than 40m as well as bare-earth DEMs were unsuccessful in predicting an accurate commshed while LiDAR resolutions coarser than 15m provided significant predictions of equal accuracy. These results indicate that high resolution LiDAR is needed to accurately model commsheds but signal strength threshold classification determines which of these higher resolutions are significant.