269

1 Introduction

Internet Protocol (IP)-based networks and

TCP/IP-applications are steadily getting more

popular. Many applications require more than

best-effort service, i.e. they need Quality of Ser-

vice (QoS) guarantees. Studies of customer

expectations [Alt00] indicate that users would

like service differentiation. Today’s networks

cannot support different requirements coming

from different applications, since the methodol-

ogy, mechanisms and their implementations are

not yet mature. Some mechanisms assuring ser-

vice differentiation and QoS support in IP-based

networks are available, like the Integrated Ser-

vices (IntServ) and the Differentiated Services

(DiffServ) standardised by IETF, but it is not

clear how to use these in order to deliver the ser-

vice end-to-end with the quality agreed with a

user. Nor is it clear which parameters to use to

express quality at the application service level,

i.e. the quality the user can directly perceive.

Moreover, mapping these parameters to network

performance parameter is not a trivial task. In

addition, no services are charged for the quality

they are provided with.

The EURESCOM project P906-GI QUASI-

MODO (Quality of Service Methodologies and

solutions within the service framework: Measur-

ing, Managing and Charging QoS) tried to

answer some of these issues. The main idea of

offering a reasonable set of quality classes to

users according to their needs and possibilities

(e.g. to pay) was investigated by developing

and implementing the QUASI-model.

2 The QUASI-model

The QUASI-model [P906-1] was intended as a

simple and practical way to offer several classes

of service to customers. As a basic assumption,

it was decided that guarantees could only be

offered within the network under provider’s

control, i.e. between certain edge routers called

Measurement Reference Points (MRPs). In order

to provide end-to-end quality to the user, the

user’s domain (e.g. LAN, CPE, applications) has

to be characterised, i.e. their contribution to the

overall quality has to be taken into account.

Therefore, in the simple scenario depicted in

Figure 1, it is assumed that the provider can

offer a service directly to the end-user and con-

trol the network portion between MRPs (A and

B). Moreover, the provider would characterise

the user’s domains in terms of describing mini-

mum system characteristics and performance

requirements (points C – A; B – D). The users

are connected to the provider’s network through

an access network – only the LAN access was

investigated in the project, but with modifica-

tions the QUASI-model can be applicable to

other access networks. The users are assumed

not to be very literate in technology and QoS in

particular, implying they are not supposed to set

any mechanisms themselves. On the contrary,

they should get QoS as offered from the opera-

tor, who can only guarantee quality between the

MRPs.

During the project it was understood that this

service cannot be offered directly to a user, since

the guarantees were expressed in terms of Net-

work Performance Level (NPL) parameters, i.e.

delay, jitter and loss. Therefore, a QUASI-aware

business model with a role of a Service Provider

(SP) added, was introduced.

A business model, in general, describes different

roles involved in service provisioning, and their

corresponding relations. By role is assumed a set

of activities a business organisation (or an actor)

can perform in order to produce/consume a ser-

vice. Different roles exchange information and

have relationships. Some examples of roles are:

user, customer, vendor, service provider, net-

Methods for Monitoring, Controlling and

Charging QoS in IP Networks

JORMA JORMAKKA, IRENA GRGIC AND VASILIOS SIRIS

Figure 1 Original QUASI-model physical scenario

Irena Grgic (30) received her
BScEE and MScEE from the
Faculty of Electrical Engineering
and Computing at the University
of Zagreb, Croatia, in 1996 and
1999, respectively. She is cur-
rently working at Telenor R&D
as a Research Scientist within
the Future com Business pro-
gramme. Her main areas of
interest include Quality of Ser-
vice issues, Service Level
Agreements handling, manage-
ment, negotiations, as well as
pricing and charging for IP-
based services provided end-
to-end in a multiprovider envi-
ronment. She has participated
in several international projects
handling these topics, among
them EURESCOM P906-GI.

[email protected]

Jorma Jormakka (45) is a Pro-
fessor in the Networking Lab-
oratory of Helsinki University of
Technology (HUT) and in the
Technical Institute of the
National Defense College of
Finland. His current research
focuses on protocols, communi-
cation software, services, secu-
rity and QoS issues. He holds
a PhD in mathematics from the
University of Helsinki and an
MSc in Electronics Engineering
from HUT. He was the task
leader of Task 4 in EURESCOM
P906-GI.

[email protected]

User

Network

Operator(s)

Network

User

Network

Characterized Provided Characterized

AB

C D

Telektronikk 2/3.2001