Chris Anderson Presentation at the U

Chris R. Anderson

Prof. Chris R. Anderson will present a seminar titled "Integrated Terrain and Clutter Propagation Modeling for 1.7 GHz Spectrum-sharing" on Friday 26 January 2018, from 3:05 - 3:55 PM in WEB 1230.

Abstract: In 2015, the FCC conducted the Advanced Wireless Service 3 (AWS-3) auction, which sold 50MHz of spectrum from 1695-2180 MHz for commercial use. This spectrum was auctioned to meet the ever-increasing demand of spectrum from mobile LTE users. This spectrum, however, overlaps with dozens of military communication systems, including over 40 Navy air-to-ground systems. Although these systems are to be transitioned out of the band, many will not fully depart until 2025 and their operation would be degraded by the presence of in-band interference from LTE networks. To support commercial operators during the transition period, an Early Entry Portal Analysis Capability (EEPAC) was created to provide a mechanism for commercial operators to begin deploying their networks if propagation modeling and aggregate interference analysis could demonstrate that their network would not cause harmful interference to DoD systems. The initial analysis setup, however, proposed highly conservative model parameters in an effort to ensure protection of DoD systems. Despite ongoing measurement campaigns that support less conservative assumptions, many questions remain concerning the dominant propagation mechanisms, clutter effects, and aggregate interference modeling between DoD and commercial users.

This talk presents the results of a narrowband measurement and modeling campaign for the 1.7 GHz band. The objective was to fully characterize the large-scale propagation in these bands in an effort to update current predictive propagation models and techniques. Over 350,000 received signal strength measurements were recorded in eight different geographic regions of the U. S. Measurement results were compared against empirical and semi-empirical models, including Log-Distance, the Alpha/Beta/Gamma, Specific Attenuation, and the Irregular Terrain Model (Longley-Rice). Results indicated that the classical models provided a very poor fit to the measured dataset, with mean and RMS errors in the range of 15-30 dB. A simple Attenuation Factor model using publicly available Geographic Information Systems (GIS) datasets, however, was demonstrated to significantly outperform all of these models, suggesting that significant revision to AWS-3 propagation modeling should be undertaken.