Abstract

We compare LTE coverage data from the FCC with a crowdsourced dataset from Skyhook for New Mexico. While the two coverage datasets tend to agree in urban areas, there is significant disparity, up to 15%, in rural and tribal census blocks. On-ground LTE coverage measurements collected across 120 miles of rural and tribal New Mexico indicate that even the crowdsourced data exhibits over-reporting, although to a lesser degree than the FCC data. The findings make a case for including mechanisms to validate ISP-reported FCC coverage data. While crowdsourcing is a good alternative, targeted active measurement campaigns are needed in areas where existing crowdsource datasets are sparse.