5 traffic forecasts
TRANSCRIPT
-
8/9/2019 5 Traffic Forecasts
1/3
Traffic Forecasts: Telephony
Provide a basis for decisions on investments in the network; should give
information about future demands for telecommunications.
Traffic forecast must be accurate as possible and should be derived from
present traffic figures which are as accurate as possible Importance of these forecasts is increasing; market plan and the product plan
rely heavily on them.
Time aspect is very important since its only an estimate; network operators
therefore use forecast periods of different lengths
As a rule, the following activities are included
Forecasting methods
!orecasting method determines basic data to be used and, to some e"tent, the
outcome.
Three #$% groups of methods
Purpose and conditions defined
Traffic measurement data, data of population growth and
economic development, and outcome of earlier forecasts.
&e'uirements for data accuracy and the forecast period.
(ata gathered is processed.
-
8/9/2019 5 Traffic Forecasts
2/3
). trend methods;
*. statistical demand analysis; and
$. assessment methods.
Trend methods
!ind main trend over a long or short period.
+est suited for obects with even growth and for short forecast periods.
If the development is e"ponential, the trend can be appro"imated as the mean
percentage growth during the period #similar to the calculation of compound
interest%.
-"ample
If the number of subscribers in an e"change area has increased by about * in the
past five years, then this increase marks the main trend. /e can draw the conclusion
that in the ne"t three0year period there will be $ " * 1 2 new subscribers in the
area.
!igure A trend method
Statistical demand analysis
3eeks to map out those factors which underlie the earlier development.
4hanges that can be e"pected during the forecasting period are then collated.
-"ample
-
8/9/2019 5 Traffic Forecasts
3/3
A town may be dominated by one big company that has recruited about ) people
annually in the last few years. If this need for recruiting continues to e"ist, one may
reckon that the growth ascribable to the company will keep on at the same rate.
/ith these background factors as a basis, the network operator creates mathematical
models describing the current conditions. The relationships are then used forforecasting.
+asic data is obtained from statistics processed by labour market authorities, county
or city administrative boards and other bodies.
Assessment methods
5se of assessment methods re'uires that a forecast should be based on
previous e"perience or collected information.
+asic data is not systematically analysed. These models can be used when
preconditions for trend proection or statistical demand analysis are lacking.
6ne variant is known as the analogy method it is assumed that situations or
obects with similar preconditions will develop similarly.