|
|
| |
 |
 |
|
 |
| Today’s
wireless networks are constrained by the physical
realities of radio frequency (RF) propagation
and the need for an operator to take a limited
amount of RF spectrum and continuously reuse
it to achieve market-wide coverage for their
wireless subscribers.
With frequency reuse, where a frequency in a
given geographical area is also used in another
area, base stations interfere with each other.
This is inherent in cellular theory. Frequency
reuse is essentially unavoidable since wireless
operators are given a scarce resource (spectrum)
to supply services to wide areas. However, when
this interference (called co-channel interference)
becomes severe, networks suffer diminished quality
of service and inefficient spectrum usage. This
issue becomes is becoming more pronounced as
demand for additional network capacity grows.
This is because increased capacity in cellular
networks is typically achieved by “splitting”
cells (reusing frequency more tightly within
a geographic area). The additional capacity
and signal quality requirements of data services
will also contribute to growing interference
problems in cellular networks. |
Historical
Perspective
The complexity of wireless
networks has increased many-fold since commercial
“cellular” wireless services were
introduced roughly two decades ago. Back then,
operators were largely designing and managing
a voice-oriented, analog modulation system that
involved cells that were typically several kilometers
in diameter. Today’s networks are very
dynamic, often involving multiple air interface
standards for voice and various data rates,
each with different performance requirements.
Even tough some of the RF planning principles
are similar; the interference tolerance of today’s
networks is much lower. Interference, for example,
greatly limits data throughput in GPRS and Edge.
Much traffic is supported by micro and pico
cells, where coverage is less predictable than
coverage in macrocells unless spectrum is segregated
with the resulting overall less efficient use
of spectrum. All systems employ digital modulation,
which is characterized by rapid service degradation
beyond the interference and noise thresholds.
In the interference-limited environments of
most modern systems, operators must carefully
control the “footprints” of cells
in order to maintain service quality targets
while maximizing spectrum reuse. |
RF
system design has traditionally relied heavily
on the use of propagation models to predict
the coverage provided by cells. Many model refinements
have evolved to support analysis of more complex
systems, including development of more sophisticated
propagation models and the use of more accurate
terrain, building andmorphology databases. However,
all propagation models have prediction errors,
at best averaging error of a few dB over a cell’s
coverage area, and often with prediction errors
approaching 8 dB (it follows that such error
can be of up to 16 dB in C/I ratio calculation).
Such errors,when used to analyze systems with
Carrier to Interference ratio (C/I) requirements
in the range of 10 dB, provide a severe limitation
on the accuracy of performance predictions based
upon propagation models, and limit engineers’
abilities to maximize actual network performance.
The same problems apply to the ever more used
Automated Frequency Planning (AFP) tools. |
Field
engineers have a variety of RF performance measurement
systems available to them. The technology used
in these systems falls into two general classes;
i) test phone systems which have special software
that supports a technical analysis of network
behavior, and ii) scanners which can monitor
the signal strength and overhead channel information
of scanned RF channels. Test phone systems provide
valuable data that reflects the actual subscriber
experience, and, like the subscribers’
experience, repeated tests at the same location
may yield different results depending upon activity
levels in the network. Scanner data is more
constant, as the channels measured are typically
those that are timeinvariant. Both types of
systems are largely limited to providing analysis
of signals from the dominant cell on a particular
RF channel – the presence of interfering
channels can be detected, but no further information
can be gathered regarding the source or magnitude
of the interfering channel. |
Click here to find out
more |
|
|
|
|
