Abstract: This report presents spectrum occupancy data of the 3550–3650 megahertz (MHz) maritime radar band measured in June 2012 near San Diego. In this band, the military operates SPN-43 air marshaling radar systems with well-defined signal characteristics. A measurement system and frequency-swept technique were designed specifically to detect SPN-43 emissions. Over the two-week measurement duration in June 2012, we observed multiple systems operating simultaneously in band, spectral spreading of SPN-43 emissions, and out-of-band pulsed emissions that spanned the entire band of interest. In this presumably high-usage mostly military spectrum environment and at a low occupancy threshold (i.e., -83 dBm in a 1 MHz bandwidth at the output of a 2 dBi antenna), mean band occupancy was {36.6, 7.5}% during {weekdays, weekends}. There was a {40.0, 59.8}% chance that the band was empty and a {18.4, 2.3}% chance that the band was full. During weekdays, spectrum usage was dominated by out-of-band pulsed transmissions that spanned the entire band at a relatively low level (approximately 10 dB above the low threshold level). These signals were superimposed on the stronger SPN-43 signals. On average during weekdays, transmissions occurred every four hours with a mean transmission length of approximately 1.5 hours. In contrast, weekend spectrum usage was primarily SPN-43 transmissions at 3.59 GHz arriving every 29 hours (on average) with a mean transmission length of approximately 9 hours. Measured SPN-43 signal amplitudes were at times strong enough to overload the measurement system (exceeding the -83 dBm threshold level by more than 60 dB). Statistical considerations that arise when measured data are used to characterize spectrum occupancy are also discussed in this report. Covered topics include channel occupancy definition, estimation, and uncertainty.
Keywords: spectrum occupancy, maritime radar, spectrum management
The full report is available at http://www.its.bldrdoc.gov/publications/2747.aspx