decaturtxcowboy wrote:

> Tower had a 1900 MHz AT&T at top and 800 MHz Cingular right below
> it. We took two Nokia 61xx phones running in field diagnostics mode
> which displays signal level in one dB increments. In a rural
> environment and walking into small business, open road in a truck,
> on foot...both phones had similar coverage.

That's a tough comparison to make accurately both because of the
absolute accuracy of the RSSI built into the phones and because of the
variability due to propagation. Still, an interesting one to attempt!

Further on the topic of 850/1900 frequency dependent propagation
differences...


At the risk of hearing more "nonsense" responses I'd like to submit that
there truly is a frequency term in outdoor propagation models and that
for typical scenarios, there indeed is a significant difference between
850 MHz and 1900 MHz, attributable to the difference of transmission
through real impairments, like buildings and foliage as well as to the
differences in diffraction between the different wavelengths. The
article that John Navas suggested as relevant:

http://www.sss-mag.com/pdf/1propagation.pdf

is mostly discussing frequency selective fading and multi-path
impairments rather than mechanisms for attenuation. Slide 8 does show
curves for urban, suburban, rural and LOS but there is no detail
showing the derivation of the model and is not relevant to this discussion.

For a relevant comparison, I would suggest the COST231 propagation model
with Hata extensions. COST231/Hata applies to path lengths generally of
1 km or longer and beyond the LOS region discussed in the paper above.
The Hata extensions extend the useful frequency range of the model. One
can Google for Matlab implementations of the model if desired.

Applying this model to 850 and 1900 MHz scenarios, one really does see a
significant difference in attenuation due only to frequency. The precise
difference depends upon the transmitter and receiver antenna heights
selected and on the particular subset of the model. However, differences
in the area of 6-12 dB region are common. As an example, for 20
meter/1 meter antenna heights in a suburban environment, at 1 km the
model gives

radiosuburban_atten(850e6,1000) = -107.37
radiosuburban_atten(1900e6,1000) = -114.42

for a difference of about 7 dB.

If antenna/frequency effects can be avoided (something not necessarily
done in your walk-around test, by the way), by using constant physical
aperture on one antenna and constant electrical aperture on the second.
Doing this assures constant ERP and constant receive aperture so that
wavelength falls out of the freespace portion of the result. What
remains shows the frequency dependent effect in the model.

While one can argue which model is most correct or useful, the
experiential result is still a very significant difference between
propagation in the two bands in typical environments.

The above example is just one scenario but were complete coverage
required, the 1900 MHz scenario would require reducing the maximum radio
path by more than a factor of two. As a result of that reduction, over
an area the number of base stations would have to more than quadruple,
for equivalent coverage.

g