TECH TIPS
Updated 07/05/05
We will periodically offer
tips that we hope are of real use to technology and engineering professionals
in the wireless industry.
If you
have a tip you would like to share, or have a technical question that you think
would lead to one, please send it to techtips@drucker-associates.com.
Be sure to
check back occasionally for new Techtips.
07/05/05 Performance testing of CDMA handsets in live networks
All wireless networks are to
a certain degree dynamic in nature. But
channel conditions in CDMA systems, including both CDMA-2000 and W-CDMA, are particularly
sensitive to channel loading and other operational variables. It is therefore quite difficult to obtain
meaningful quantitative results for performance testing of compatible handsets
in “live” CDMA networks – a call that progresses flawlessly over a particular
drive test route one day might be dropped the next.
For a more useful
alternative to pure quantitative testing of handsets, operators might want to
consider benchmark testing in which performance of the handset under test is
compared, side-by-side, with that of a “gold standard” model.
07/20/04 Call Erlangs vs. code Erlangs
In CDMA networks, the total
number of Walsh code Erlangs delivered in a given hour will always far exceed
the total traffic Erlangs delivered in the same hour. This is because of soft/softer handoff – at any given time, a
call may be served by several sectors, each of which devotes a Walsh code. Understanding the difference between call
(or traffic) Erlangs and code Erlangs is critical to evaluation and optimization
of capacity and efficiency for each sector in the network.
06/10/04 An interesting CNN story
CNN recently ran a report
on the state of wireless service in the US.
The picture is not pretty.
Quoting a University of Michigan study of 40 service industries,
wireless ranks next to last in user satisfaction, beating out only cable television. According to the article, customer gripes
are many, but it is clear that the number one issue is quality of service. The real problem may well be that operators
have a very poor understanding of customer perception of service quality. This
is primarily because the metrics most widely used to gauge quality – dropped
call rates and the like – don’t really tell the complete story. For more on this, see a recent Wireless
Week article by Elliott Drucker.
04/15/04 More on drive testing
The previous Techtip
(below) stresses the need to take measurements for inter-cell interference, and
potential interference, in drive testing.
Generally, that requires, for each geographic data point, measuring path
loss from/to each nearby sector. This
is most easily done by measuring the received signal strength from a
transmitter of known power level on a discrete channel. A few points are worth noting.
First, the number of sectors
that should be measured at each location is usually much higher than one might
imagine. The point is to measure path
loss from each sector that might contribute to interference, which means
all sectors with path loss less than maybe 15 dB greater than the path loss
from the serving sector. In urban
areas, this can often require measurements from 20 or more sectors at any given
location.
In CDMA networks, it will be
difficult to observe pilot channels that are 15 dB weaker than the strongest
pilot. Measurement of Ec/Io
on the two or three strongest pilots does provide an indication of the general
interference situation, but not complete path loss information. It also does not provide full information
about the uplink interference contributed by mobiles operating at specific
locations. Therefore, rigorous path
loss based interference drive testing really requires use of discrete
narrow-band transmitters.
Because of the difficulty in
engineering substantial numbers of discrete frequencies for drive testing,
operators should consider the alternative of a single dedicated uplink
frequency transmitted by the drive test mobile and simultaneously received at
all sectors. By combining time-stamped
RSSI data from each sector and combining it with time-stamped position
information from the mobile, highly accurate path loss measurements can be
obtained from drive testing.
03/09/04 Drive test for interference, not just coverage
Just about everyone has seen
the Verizon Wireless ads with the little guy in the jumpsuit who roams the
wireless world forever asking, “Can you hear me now?” In terms of pointing out how crucial empirical testing is to
verifying wireless network performance, the message is right on. But as a demonstration of how such testing
should be done, the spots leave a whole lot to be desired.
The point is, if you are
going to undertake the substantial task of extensive drive (or walk) testing,
you want to get more out of it than simply a map of where coverage is
“adequate”. In urban networks,
engineering and optimization, particularly with respect to capacity, is driven
mainly by the need to manage interference.
But simply measuring the receive level of the strongest pilot or
downlink control channel, which is too often the case for engineering drive testing,
does little to improve knowledge about interference. To obtain this much more valuable data, the mobile test system
needs to also record signal levels from every potentially interfering
sector. This requires much more
sophisticated equipment and measurement processes, but remember that the
biggest cost in empirical testing is the labor-intensive data collection. Spending more on the equipment used in order
to get a big boost in the value of the output thus makes a lot of sense.
How is your network
performing?
Most wireless networks
operate at well below potential capacity, and often deliver less than ideal
service quality. Elliott Drucker,
President of Drucker Associates, is one of the industry’s leading authorities
on wireless network optimization. If
you are concerned about how your network is performing, you should give Elliott
a call to discuss strategies for maximizing your spectrum efficiency. He can be reached at (+1) 425-820-3411. Or e-mail him at edrucker@drucker-associates.com.
Issues with network
security or intersystem /inter-network functions?
Talk to our associate
David Crowe, President of Cellular Networking Perspectives, Ltd. David is a leading authority on the critical
issues of intersystem/inter-network interfaces and network security. Visit the CNP website at http://www.cnp-wireless.com.
Drucker Associates Home Page