[By Subject] [By Date] [By Sender [By Thread]
Previous In Time

Next in Time

Phil Karn :
Fri, 6 Dec 1996 16:12:20 -0800 (PST)
Much useful numerical data is available in the results of our CDMA field tests in 1991-1992. Some of this is on the web; see . One particularly useful piece of info is the distribution of mobile transmitter power levels (because of automatic power control, this can varies rapidly). The power ranges from nanowatts on the low end up to 300 mW on the high end (the peak power capability of the mobile), but the distribution peaks at about 5 mW (yes, milliwatts). That's for a typical cellular mag-mount antenna stuck on the top of a car talking to typical cell sites over typical paths of a few km. The cell sites were on the tops of buildings, on phone poles and in one case on a freestanding tower. The data rate to which the power measurements correspond was about 8kb/s; if the system were to be scaled to a higher or lower data rate, the power would scale proportionately to keep a constant energy per user data bit.

The frequency band for the tests was the existing UHF cellular band at 800 MHz. The results can be extrapolated both up and down in frequency according to fairly well understood theory; the main change is the fading rate as a function of vehicle speed, which affects the effectiveness of coding and transmitter power control. We currently have systems operating at both 800 and 1.9 GHz. I don't know of any experience on bands well below 800 MHz, but I think we can make some reasonable extrapolations.

A useful experiment for our purposes would be to measure the typical path loss, antenna connector to antenna connector, between a typical VHF/UHF weak signal station and either another fixed amateur station or a mobile station. If these measurements are made with narrowband signals, they should be repeated at a variety of frequencies and averaged to take out frequency-selective multipath effects.

Along with a measurement of N0 (thermal and background noise spectral density) at the receiving station, we could then compute quite easily the transmit power spectral density into the "interfering" antenna that would equal the receiving station's N0. This would give us some hard figures that we could compare to theory. Any volunteers?

Phil ------ Submissions: vhf@w6yx.stanford.edu Subscription/removal requests: vhf-request@w6yx.stanford.edu Human list administrator: vhf-approval@w6yx.stanford.edu