R-6602(Carey) Service Contours a relatively simple models that have historically been used to define the "Service Area" of
a transmitter for the purposes of frequency coordination studies and in the case of commercial services such a broadcasting, for
legal definition of the area served by the station.
The model is based on the ERP (effective radiated power) of the station, and it's HAAT (height above average terrain) using the FCC
2-10 mile, eight radial method described in FCC Part 73, and FCC Part 22. The general definition of a "service contour" is that inside
the contour, the signal strength will meet or exceed a certain value 90% of the time (time variability) in 90% of the locations (space
variability).
Field strength values expressed in dBu (decibels above 1 uV/meter) are calculated for each amateur band based on the following:
A usable receiver sensitivity is computed using and IF bandwidth of 16kHz, and RX NF of 7 dB, and a desired C/N ratio of 18 dB. This
yields a usable RX sensitivity (not squelch threshold) of about -107 dBm (1.0 uV) and is typical for many mobile radios.
A relabilty correction factor of 12 dB is then added due to the fact that Carey models use F(50,50) log normal distributions. To
increase the probability to 90% a 12 dB factor is added to the -107 dBm value. This value of -95 dBm, guarantees that at a mobile station
experiencing multipath fading, will at least 90% of the time in 90% of the locations have a value of signal that exceeds -107 dBm (1.0 uv).
Had we not made such a correction the time and location percentage would be the median value (50%), which is of no interest for service
contour applications.
Finally this value is converted for each amateur band, to a Field Strength Value (in dBu). Note that for a fixed input power to a
receiver and a fixed antenna gain, the Field Strength required to produce that input power increases with frequency. Thus for each
amateur band (10M thru 13cm) the Service Contour Field Strength values increase.
Band Value
10M 9 dBu
6M 14 dBu
2M 23 dBu
125cm 27 dBu
70cm 33 dBu
33cm 39 dBu
23cm 42 dBu
13cm 47 dBu
R-6602 (Carey) Interference Contours are constructed in a similar manner to Service Contours. Inputs to the algorithm is simply ERP
and HAAT as was the service contrours. The interference contour represents the area inside which the field strength can be expected to
be at least the controur value at 50% of the locations for 10% of the time.
Historically a different set of curves were used to determine the service contours. Both Service and Interference contours techniques
were based on propagation measurements that were made as far back as 1947. At that time lacking any sort of detailed terrain database and
due to the statistical nature of radio propagation (especially tropospheric variations that occur over relatively long times), service and
interference curves for many years provided about the only objective criteria for comparing or determing base station coverage.
The corresponding values that SERA uses for Interference Contour Field Strength is as follows:
Band Value
10M 3 dBu
6M 8 dBu
2M 17 dBu
125cm 21 dBu
70cm 27 dBu
33cm 33 dBu
23cm 36 dBu
13cm 41 dBu
With the advent of computers and digital terrain databases, it became possible to evaluate transmitter coverage with a great deal of
accuracy with various propagation conditions. While there are a number of propagation models to choose from, Longley Rice appears to
be the best suited for amateur radio repeater and auxiliary station analysis. While it's computationally more complex than most models,
a modern day PC can handle the computations in a matter of minutes for a typical system.
Longley Rice works by creating a matrix out of the area surrounding the base station, generating terrain data for each "cell" in the
terrain matrix, and then calculating the path loss between the matrix cell and the base station antenna. Then based on the ERP of the
station in that azimuth, the field strength's at each point in the matrix can be calculated. The Longley Rice model takes into account not
only terrain along each matrix cell path, but also earth dielectric constant, soil conductivity, ground cover (forest, urban, over water
paths, etc.). For fixed K factors (earth curvature), Longley Rice simulations normally have standard deviations of about 5 dB compared with
actual drive tested system. This is considered good as it's not uncommon for the drive system measurement equipment to have standard deviations
of about 2 dB.