There are other similar observations in a multi-
operator situation. If there is an operator offering
predictable prices, while some other offer load
or volume dependent prices, there may be a pro-
tection strategy for the ISP. This means that it
can offer fair prices (in the sense of a martingale
measure) to end-users and have the strategy of
shifting the traffic to the predicable prices in
case lower usage-based prices are not offered
by other operators.
Another observation is that there are cases when
changes in traffic volume are predictable, like
the traffic variation during the day to a large
extent is. If volume changes are predictable and
there are operators offering flat prices and others
offering usage based prices, the ISP may shift
traffic on high traffic times to the flat rate and on
low traffic times to the usage based rate. These
may influence the viability of a particular pric-
ing scheme, and there are also countermeasures.
In this example the operator may require that
capacity is bought only for time units of a cer-
tain duration, that predictability of traffic cannot
be used by an ISP for unwanted optimising pur-
poses.
This paper described some results of the
QUASIMODO-project, with emphasis on the
parts where the authors of the paper worked. The
QUASI-model implementations for measure-
ment and management contain measurements for
QoS monitoring, mapping users’ (QC, AC)
selections to DiffServ and IntServ architectures,
and some additional QoS management methods.
The charging-related work in the QUASI-
MODO-project included both generic charging
and accounting model, as well as two possible
billing system solutions. Here, the theoretical
findings have been stressed, while more details
on the particular implementations can be found
in [P906-7], [P906-8]. The linking between the
measurement and management implementations
and the billing systems was not demonstrated; it
is realised by the possibility of obtaining NPL-
measurements to the charging system.
Acknowledgement
Although this paper represents the view of the
authors, this way we would like to thank all the
P906-GI participants for the interesting work
and discussions undertaken. A special thanks
goes to Mr. Imad Ossaily and Mr. Denis Karpov,
who worked with the QUASI IntServ implemen-
tation. Several figures are from the QUASI-
MODO deliverables.
11 References
[Alt00] Altman, J, Rupp, B, Varaiya, P. 2000.
Quality Matters: Some remarks on Internet Ser-
vice Provisioning and Tariff design.
Telektronikk, 96 (2), 20–25.[CA$hMAN] Songhurst, D J (ed). 1999. Charg-
ing Communication Networks: From Theory to
Practice.Amsterdam, Elsevier. ISBN 0-444-
50275-0.[P906-1] EURESCOM P906-QUASIMODO:
Deliverable 1 – Offering Quality Classes to end
users.Heidelberg, EURESCOM, May 2000.[P906-2] EURESCOM P906-QUASIMODO:
Project Report 2 – Methodologies and tools for
QoS measurement and management.Heidelberg,
EURESCOM, December 2000.[P906-3] EURESCOM P906-QUASIMODO
Technical Information 1 – Survey of existing
methodologies and tools for QoS measurement
and management.Heidelberg, EURESCOM,
December 2000.[P906-4] EURESCOM P906-QUASIMODO
Technical Information 2: QUASI-model Imple-
mentation.Heidelberg, EURESCOM, December
2000.[P906-5] EURESCOM P906-QUASIMODO
Technical Information 3: Experimental evalua-
tion of QoS measurement and management.Hei-
delberg, EURESCOM, December 2000.[P906-6] EURESCOM P906-QUASIMODO
Technical Information 4 – Key QoS charging
issues.Heidelberg, EURESCOM, 2001.[P906-7] EURESCOM P906-QUASIMODO
Deliverable 3 – Methodologies and policies for
QoS charging.Heidelberg, EURESCOM, Febru-
ary, 2001.[P906-8] EURESCOM P906-QUASIMODO
Technical Information 6 – Summary of QUASI-
MODO Findings on QoS.EURESCOM, Heidel-
berg, February 2001.[SoKe97] Songhurst, D, Kelly, F. 1997. Charg-
ing Schemes for Multiservice Networks. In:
Proc. ITC-15. Ramaswami, V, Wirth, P E (eds).
Amsterdam, Elsevier.http://www.eurescom.de/Public/Projects/p900-
series/P906/P906.htm