A BUSINESS APPROACH TO
NETWORK DESIGN
September 2002
A Hard Nosed World It used to be easy to access
investment capital for telecom projects and extensions.
However, anyone
who has tried, in the last year or so, to get financial approval
for network extensions knows, that the world is very different
now. Money is extremely tight. Vague promises of “higher
quality” and the like are no longer good enough. Higher
profitability is the only aim. All investments in infrastructure,
therefore, have to be justified in terms of hard finance.
That is not always easy to do.
Network Financial Modelling
Increasingly attention is being focussed on network financial
modelling.
This works by looking at the costs of Interconnect. Specifically
the amount that you pay to Interconnect partners to terminate
calls from your subscribers.
Simply the idea is to calculate the costs of Interconnect using
the actual operational network today and forecast the likely
future costs using trend analysis. Next you calculate what
the costs would be once you have implemented a proposed network
extension and forecast those likely costs using trend analysis.
It is then a simple case of comparing the accumulated difference
over your required payback period against the cost of the proposed
network extension.
Checking Other Options
The first step is to check out alternative ways to reducing
Interconnect costs. It may be that considerable savings could
be made without implementing network extensions. These need
to be considered in order to generate a genuine base line.
This involves an analysis of the potential rates that you might
use from existing and potential partners. These need to be
analysed against your actual call patterns in order to generate
a lowest possible, as opposed to actual, cost.
Optimum Route
This least cost route now needs careful examination for each
significant destination. This is done using the current interconnect
points that are currently in place for your network. It needs
to be compared with an actual routing report.
The ideal least cost route my actually be constantly overflowing
onto a default, more expensive route. It might be the least
cost route but not the optimum route to select for your purposes.
The optimum route is the route that minimises the cost whilst
at the same time provided proven capacity in the network routing
required to terminate your traffic.
This can be thought of as the optimum “network constrained” route
for each destination.
Without Constraints
The next step is to run the exercise again but without any
network constraints in place.
This will show up a series of potential savings that might
flow from network changes.
The significant potential savings, those that show the most
promise, now need to be investigated in greater depth. The
potential changes should now be “assumed” in
the network model, one by one, and the exercise run again.
Considerations
of capacity, projected capacity, quality of service and so
on also need to be made to ensure that the savings can be
realised in practice. Future trends also need to be estimated
for traffic.
This should therefore yield a figure for the projected savings
for an individual network extension over your standard investment “payback
period”.
The Challenge Of Complexity
This is a complex technique that uses a lot of external data
sources such as tariff rates from potential partners and compares
them against existing and projected traffic patterns. It has
in use, however, pointed up changes that have meant significant
reductions in termination costs for a number of operators.
The complexity forms a real challenge however. The greatest
problem lies in the repeated rating of actual call records
against a variety of proposed network models and potential
Interconnect partners.
For a network of any size this can prove to be a daunting prospect
unless the data is organised correctly to begin with.
Aggregation
The best way of organising call data records is to use a process
of aggregation. This means accepting raw CDRs, before carrying
out any rating, and storing them in aggregates that share common
characteristics. These characteristics can include such variables
as:
Billing Operator
Call Destination
Product
Minutes
Day call occurred
Timeband
Charge-type (CR / DR)
Billing remarks
Discounting properties
Switch
Trunk-Group ingress and egress
Point of Interconnect;
Physical call direction If the CDRs are totalised within these
aggregates then it becomes much easier to manipulate the date
store.
The next issue is to find a way of making the loading of rates
from potential operators as quick and as easy as possible.
These need to be accepted and translated into a common format
so that they can be applied sequentially and in an automated
way to the CDR aggregates.
This will allow multiple rating and re-rating of the CDRs
to allow for the various “what if” scenarios
that need to be evaluated.
Results
This exercise seems complex. However it is the only way to
continually tune the network both in terms of its physical
links and in terms of the Interconnect partners selected.
Recent surveys have suggested that, for most operators, the
cost of Interconnect termination approaches 55% of the total
ongoing costs of running their business. It is not, therefore,
something that should be ignored. In manufacturing industry
most companies have a Director of Purchasing who is continually
looking at the cost base of the company to see how it can be
improved. His performance has as much of an impact on the company
as that of the sales team. Relatively small savings on the
costs of raw materials or components going into the factory
can make a huge difference. In established and mature markets
the control of costs is a key strategic focus. Indeed for many
companies it has become regarded as the key fundamental axis
of competition. Long term studies have shown that the lowest
costs tend to go with market leadership and in turn with the
highest levels of profitability.
In telecom companies, for a long time, the focus has instead
been on gaining market share, on the acquisition of new customers.
We have traditionally not had quite so much of a focus on the
cost base. With the maturing of the telecom market, a continual
financial analysis of the network, and its capacity for improvement,
may well become a pre-requisite for survival. Certainly for
those operators who only deal in raw bandwidth without any
added value services it may be the only way to compete. |
 |