Question 14/II - Methods for forecasting international traffic
(continuation of Question 28/II, 19851988
study period, amended)

Considering

(a) that a number of new services have been defined which have not
been taken into account before for international traffic forecasting
purposes;

(b) that the continuing evolution of the traditional network to the ISDN
is introducing new services and enhanced services;

(c) that various services on ISDN will have different characteristics
(international time, holding time, profile, etc.) and that the
forecasting
models must take the characteristics into account;

(d) that a new data base and also new explanatory variables are
necessary for forecasting traffic on ISDN;

(e) that the integration of forecasts for traditional telephone traffic
and
other existing and totally new services requires further exploration on
a
continuing basis;

(f) that the new Recommendation E.508 is an introduction to new
service forecasting and has left several areas open for further study;

(g) that the introduction of new design methods induced by advanced
routing techniques, e.g., dynamic routing (Recommendation E.170)
implies a need for short-term traffic forecasts on an hourly and daily
basis,

1. What services will have a significant impact on the traffic on
existing
networks and ISDN?

2. How should these services be classified recognizing that a different
type of service may require different forecasting approaches?

3. How should a forecasting model for the mixed traffic on ISDN be
developed?

4. What traffic data and type of explanatory variables are required for
forecasting these services?

5. How does the quality of the data base affect the choice of
forecasting technique?

6. How can the results of market research surveys be incorporated in
the forecasting methodology?

7. What qualitative forecasting techniques are appropriate to new
service forecasting and how can these be incorporated into existing
methods?

8. What type of forecasting methods can be used for hourly and daily
traffic forecasts as a basis for advanced routing techniques?