Page 226 - GuideFWA

• The Ericsson contribution considers the effect of a single FWA base site on a nearby office
building, and subtracts the channel usage (in Erlangs) of the FWA site from the total
available. The contribution claims that the remaining capacity is available to UPCS systems,
so that coexistence is not a problem.

• The Lucent contribution considers the aggregate interference from multiple FWA sites and
computes the probability distribution of the interference power per channel, and also for the
least-interfered channel. This calculation yields the effective increase in the noise floor as
seen by the UPCS system.

Comments were made on both contributions by members of the Experts Group.

Comments on the Lucent contribution included:

• A macrobase (sectored) FWA base model should be used, rather than microbases.
• The assumed FWA Erlang density was too high.
• The UPCS system can accommodate the FWA interference by making the UPCS cells

smaller.
Comments on the Ericsson contribution included:

• The analysis was over-simplified and misleading.
• The interference from the nearby FWA site is irrelevant, since those channels would

never be selected by the UPCS system.
• It ignores the noise floor increase on “available” channels used by further-away FWA

sites.

5.1.2.3 Comments on the Lucent Contribution Addressed

5.1.2.3.1 FWA Macrobase Model
The model used in the Lucent contribution can easily be applied to an FWA “macrobase”
scenario. The FWA base station uses six overlapping sectors, each with a beamwidth of roughly
90°, providing some degree of redundancy. Each sector corresponds to a single radio (12
timeslots, for DECT). With this configuration, a given face of an office building would “see”
one out of every six transmissions, on average.

Figure 5.6 shows the probability distribution of the interference on an arbitrary channel
(frequency/timeslot), and on the least-interfered channel.

212