Page 233 - Contributed Paper Session (CPS) - Volume 6

CPS1907 Klaudia M. T. et al.
developed infrastructure, standards, changes in market structure, and so one.
From the eighties to the middle of second decade of 21th century had
developed the automatization, this period is called the third industrial
revolution. The automatized technologies in industry resulted a cross-fertilized
development in computer networks, telecommunication, mobile internet and
industrial communication. The process influenced the business models, the
globalization, as well, and it was a technological driving force in development
of consumer electronics.
Since the second decade of 21th century the new technologies have
spread in fields of cyber physical systems, the internet of things and developed
networks. This radical new technology in industry is called industry 4.0, and it
requires real time solutions in telecommunication systems. (Wollshlaeger et al.
(2017))
The new technologies penetrate usually along an s-curve (logistic curve).
The innovation spreads out at a relatively low speed, because it needs time to
develop the new technology, and the enterprises, markets have to adopt to
the new circumstances. However, the adjustment results increasing
penetration rate of the new technology, until the all opportunities will be
utilized. In this case the limit of growth is reached, and the increase of profit
can be realized by adaptation of other new technology. The time between
technological changes is getting shorter. (Brynjolfsson & McAfee, (2014))
Due to the above mentioned processes the most physical indicators of
telecommunication networks have explanatory power only in a short period
until the growth of their spread reaches the inflection point. To all sub periods
can be specified one or two variables that had a significant impact on the
growth of gross value added. Because of the change of the explanatory power,
time varying coefficient model is selected to handle the time series and to
forecast. The paper shows that this approach for estimation has a more
accurate predictive ability compared to the previous applied model.

2. Methodology
If the explanatory variables change from period to period, the correlation
between physical indicators and gross value added of information and
communication should be calculated to detect the most suitable variables for
all sub periods. If the variables belonging to sub periods are confirmed, the
construction of time varying coefficient model (TVC) model is appropriate.
The fit of TVC model should follow the steps included in the Figure 1:

222 | I S I W S C 2 0 1 9