The structure of today’s telecommunications market is getting more complex than ever due to multiple reasons. First of all, the traditional monopolistic Public Network Operators’ (PNOs) market is moving towards an open market with many various players. Such liberalisation is followed by a strengthening of regulatory directives that support and favour the open market, e.g. Open Network Provision (ONP) directives from the European Commission (EC). Other characteristics of such market include the ever increasing number of: • Users with the demands for real-time multimedia applications that have high requirements on the quality of services used to support them. • Applications and services that are included in products delivered to customers. Their diversity and the fact that these are offered on various functional levels add to the complexity of the composition/ packaging process. • Providers playing different roles and offering a variety of services to users. • Technologies implemented in networks, as well as their combinations and hybrids. The structure and topology of networks are also getting more complex both logically and physically. The market described above is called a multi-service multi-provider market implying that several services are provided by several network infrastructures administrated by different providers. In such a market, a provider needs to co-operate with its sub-providers in order to deliver end-to-end a service with the agreed quality to the user. At the same time, the sub-providers may aim at the same market segment by offering same/similar products, i.e. they are potential competitors. One of the factors users observe and consider when differentiating providers in the above described market is the Quality of Service (QoS). It is important for users to avoid the consequences of not completing their tasks because the quality of the service they agreed and used was poor. Therefore are users willing to pay more to get some guarantees on the quality. Consequently, the QoS handling becomes very important for a provider operating in this type of a market. Handling QoS includes tasks like QoS analysis, provision, and assurance. The tasks are a challenge for the provider because they consider not only the implementation and the usage of different mechanisms for QoS support, but also the complex problem of assuring the quality of the service supporting traffic that crosses several domains outside of his administration and control. Note that though always positive for users and the welfare of a society, it may not always be fruitful for the provider to absolutely play on the QoS. Offering services with QoS guarantees makes him more visible to the users in the market. This usually results in the increased number of customers and reduced churn implying an increase of the profit. But, at the same time, he is more exposed to the observation and the criticism from e.g. regulatory bodies, customers organisations, etc. Different measurements need to be performed to report that agreed QoS was delivered. Promising, but not delivering the quality as agreed would seriously harm provider’s credibility and profit. A provider offering its services in an Internet Protocol (IP)-based environment needs, in addition, to tackle the challenge of the technology immaturity and partially-developed/incomplete mechanisms for QoS support. Not forgetting existing technologies, new ideas, approaches and paradigms need to be introduced step-wise and incrementally. Having a generic QoS framework that can be reused in any emerging situation would enable a provider playing in the multi-service multi-provider environment to respond to these challenges in an efficient, organised and cost-effective way. The generic QoS framework coheres QoS-related issues; it envelops the terminology and concepts that can be used independent of the service, technology, network architecture, business models and roles the provider is considering in the particular situation. Backward compatibility with already implemented QoS framework(s) and regulatory directives is a necessity for such a framework. Theoretical principles are of limited value if their usage in practice is not structured and verified. Hence, a methodology for applying the concepts of the framework is needed as well. Some parts of the methodology may even be automated, so that additional effectiveness is achieved. In the existing literature several approaches to the QoS framework may be found. Different standards and the results of various QoS-related projects initiated by standardisation bodies (ITU-T, ETSI, IETF, ISO/ODP), industrial fora (TMF, TINA, OMG, EURESCOM), and academic fora (universities) bring different QoS frameworks that are concerned with various aspects of the service provision and QoS support end-to-end. Few of them are composing both technical and business viewpoints and aspects in a complete picture. Even less of them are readily applicable in a multiservice multi-provider environment. These facts motivated the development of the generic QoS framework called EQoS+ framework, which is described in this thesis. It is a successor of the EURESCOM QoS (EQoS) framework and represents the continuation of the work started in the EURESCOM P806-GI project [p806-site]. As stated in [p806-d1], the EQoS framework is a generic QoS framework meaning that it is service/technology/network independent. Its usage should simplify the definition of interconnection agreements and, in particular, the specification of the QoS aspects in them. Any company can benefit from using a versatile framework repeatedly in order to avoid multiple repetitions of similar tasks (e.g. QoS specification, interface identification, measurement schemes identification, etc.). The EQoS framework addressed the challenge of QoS offer and management in a multi-provider environment by focusing on the agreement made between a user and a provider for a given service. This introduces the concept of one stop responsibility of a given actor (the primary provider) to another (the user). The multi-provisioning issues are addressed by allowing agreements to be defined recursively between any pair of actors, one in the role of a “provider”, the other in the role of a “user”. This recursive approach acknowledges that the delivery of a service depends on the successful operation of many components, particularly when working within a multiprovider environment. Though the EQoS framework offers a basic structure, several issues have been left open due to the limited resources in the project. The EQoS+ is a framework that enhances the EQoS basis. It is comprised of three elements: terminology, concepts, facilities. In the EQoS+ terminology, the EQoS basis is enlarged by introducing the definitions for several new terms; in particular those relevant for tackling business aspects of QoS. In addition to the one stop responsibility concept, three new concepts are introduced in the EQoS+ to further help establishing a common view between the provider and the user, and to enable a further decomposition of complex service provisioning/ delivery scenarios in relation to the functional layers and time: 1. Agreed QoS concept for establishing a common reference to QoS between a provider and a user; 2. Timelines and granularity concept, which structures different types of the agreements and explains the granularity and choice of the QoS parameters in their contents. The agreements are also related to different stages and instances of services as offered to users; 3. Layering and mapping concept that considers QoS mapping issues both within layered network models and in a value chain; it also relates the QoS requirements to the resource reservation. The EQoS+ facility SLA template proposes a generic structure for both an agreement in general, and its QoS-related part in particular. The informational model of the template may be implemented in a protocol enabling thus automatic negotiation/trading between two devices. Consequently, the negotiation process would be more automatic and effective. A methodology for applying the EQoS+ framework is presented in the thesis as well. It is described semiformally by using a practical object-oriented approach and the Unified Modelling Language (UML) [uml-spec] for the visualisation, rather than classical mathematical rigor. The methodology defines the information needed as input for applying the EQoS+ framework when considering QoS for a product/service; the procedure itself is described briefly leaving to the provider applying the framework an opportunity and privilege to implement his own solutions; the output is comprised of known relationships with selected sub-providers, and agreements made both with them and users. Also, the phases of a generic Service Life Cycle (SLC) are discussed and related to different QoS activities and EQoS+ elements. Finally, any new theory and methodology should be verified, and here the verification is done by analysing the applicability of the EQoS+ framework on an IP telephony service provision case. The benefit of the EQoS+ approach is that it formally accepts that the delivery of a service is depending on the successful operation of many components, and that the operation should be traceable and measurable, particularly when working within a multi-service multi-provider environment. It is also a step forward that QoS is related to the agreement rather than to the satisfaction of the user. The EQoS+ framework is a basis for tackling the QoS issues in any situation, and it has been already used in several IP- and UMTS-related projects [p906-d3], [p1103-d1], [p1103-d4], [rohn2], [jens5]. It is also important to note that while introducing new concepts, the EQoS+ framework also smoothly incorporates the key features of existing QoS frameworks. Though it offers several principles for treating QoS from a holistic perspective, a number of issues is still open for further study: • Business decisions in the process of negotiation, • Business games and optimisation related to the QoS parameters decisions, • The communication of the QoS parameters to the human end-users, e.g. by developing indexes, • More detailed consideration of the SLC phases and its relation to the organisational theory, • Consequence and risk analysis related to the QoS. Each of these elements represents a piece of the puzzle that would result in a QoS handbook applicable in multi-provider environments [espv2].