Page 217 - GuideFWA

depends on the lower tail of the CDF of I (i.e., the value of the distribution at low interference
levels). Hence, for analysis of systems using DCA, it is the lower tail of the interference CDF
that is important.

5.1.1.3 Computing the CDF of the interference

5.1.1.3.1 Monte Carlo Procedure

It is a simple matter to find the CDF of I using the Monte Carlo technique, whereby a large
number of computer-generated samples is used to infer the distribution. In this case the
procedure is:

1. Input the desired parameter values (d, rmin , rb , ? FWA , etc.) and the number of samples
desired. The more samples, the more accurate the distribution at the upper and lower tails.

2. Calculate Keff .

3. For each sample generate a Poisson-distributed random number J with a mean of Keff .

4. For each of the J interfering transmitters, generate random number representing its distance
from the UPCS receiver. If u is a random number which is uniformly-distributed between 0

( )and 1, the transformation r =ud2- r2 + r2 will yield a random number between rmin
min min

and d which is distributed according to (5.4).

5. Using (5.1) and (5.5), compute the power received from each of the J transmitters (in mW)

and sum them. The result is a single sample of the total interference received from the FWA

transmissions, on an arbitrary FWA frequency/timeslot.

6. Increment the each distribution bin which corresponds to values greater than the computed

sample.

7. Repeat steps 2-6 for a total of at least 10 000 samples and divide the final count in each

distribution bin by the total number of samples. The result is the fraction of the samples that

were less than the interference level corresponding to that bin, which is the desired

distribution.

5.1.1.3.2 Input Parameters

Using the Monte Carlo approach, the CDF of I was found, for the following parameters:
• ?FWA = 300 Erlangs/km2. This traffic density is taken from subsection 6.4.2 of ETSI

Technical Report (ETR) 310, which discusses coexistence of various potential DECT
applications, including FWA, known as radio local loop (RLL) in ETSI terminology [23].
• FWA downlink transmissions originate from a network of micro base stations with 20m
elevation, Gt = 10 dBi antenna gain, and omni-directional coverage in azimuth.
• The elevation of the UPCS receiver is 10m and its antenna gain is Gr = 3 dBi.
• The radius of interference is d = 5 km. Since this is less than rb for ht = 20 m and hr = 10
m, propagation from all transmitters was taken as free-space. Since contributions from

Inter-American Telecommunication Commission 203