Page 28 - ITU-T Focus Group Digital Financial Services – Consumer Experience and Protection

ITU-T Focus Group Digital Financial Services
Consumer Experience and Protection

2.4 Possible Solutions

Digital Financial Services are realized through utilization of basic services provided by a network. Assuming
that the reliability of DFS has to be very high, there are two basic ways to ensure this reliability.

• By default, the QoS level for these basic services needs to be very high too. This is where we are with
DFS today.
• The alternative is the use of robust end to end protocols which ensure the reliability of the actual service
even in the presence of deficiencies in the underlying functionality. An analogous example would be the
TCP protocol level which ensures lossless data transmission even in the presence of packet loss in lower
layers. This needs to be developed and standardized.
Such robustness can be described by key criteria for DFS. Topmost is, for each transaction, a clear indication
if it was successful or not, which needs to be consistent for both sides. Assume a transaction is composed of
a number of steps, each step being the exchange of a data token. If the transfer of a data token has no clear
“lost” criterion, but can take, in principle, indefinite time, a time-out needs to create a defined situation. The
essential property of robustness is that, if a data token now arrives after its time-out, the protocol needs to
ensure that this token is not causing any action any more.

With respect to practical aspects of DFS implementations, this poses some fundamental differences. When
the main goal is to introduce DFS in the near future, it needs to operate with the existing installed base of
end-user devices. This will automatically limit the spectrum of applicable methods to those which can be
supported by those devices. A possible drawback of this approach is, of course, the fact that if a technology
has been deployed and is widely used, it will – as long as is working without major problems - be difficult to
replace it even if the new technology is superior. This may be less an issue with respect to end user devices
as the penetration of smartphones will be continue to increase strongly due to their manifold advantages. It
may be the case that these retaining factors are more on the side of infrastructure, as introduction of new
technologies requires new investment which may, at least in the first years of usage, not be balanced by likewise
new opportunities to generate additional revenue.

3 Conclusions

The following conclusions are, with respect to the preceding clause, based on the assumption that necessary
DFS performance is achieved by ensuring a sufficiently high performance of the basic services used to
implement DFS. The case of using a robust end to end protocol is not treated here.

3.1 Conclusions for use case #1

Four different techniques are discussed in Annex B which might be used in conjunction with DFS offers for
use case #1.

• SMS is a store and forward service. Even if the share of short transfer times may be high in typical cases,
it cannot – without modifications – be used reliably for real-time transactions.

• DTMF has limited transfer capabilities and will most probably only be used to complement one of the
other techniques.

• IVR typically requires reasonably high listening quality which might pose a problem with feature phones
in environments with higher levels of background noise.
• USSD is a true real-time technique. However, the message transfer which could be used for DFS are not
standardized.

20