Abstract

Fixed wireless access is becoming a major 5G application. We measured received power for 72 urban fixed wireless links at 5.8 GHz over a period of 48 hours each. We find that 5-minute power averages show temporal variations requiring additional fade margins close to 6 dB at the 90th percentile. The average power temporal variation was found to increase with the excess path loss of the link. We propose a model that accurately characterizes first and second order link statistics. Spatial and frequency diversity measurements allowed us to also assess the effectiveness of the corresponding fade mitigation techniques. From them we conclude that the long-term fades appear to be consistent with time-varying multipath propagation, as opposed to shadowing.