A framework for quality of service evaluation in distributed environments

End users and service customers today want to perform Quality of Service evaluation for multimedia services and distributed applications. In this undertaking, they need assistance and appropriate procedures and tools. Unfortunately, today's Quality of Service frameworks and Quality of Service management products on the market do not help. This thesis fills the gap. It argues that it is possible to find a common simple Quality of Service evaluation procedure for end users and that a corresponding evaluation process can be automated. This thesis proposes a new Quality of Service evaluation framework that is suited to end user interests and needs, that is application-oriented and networkindependent, that is open and extensible, and that has the potential to complement today's Quality of Service frameworks. It encompasses: 1. A new 5-step Quality of Service evaluation procedure Its purpose is to identify those few quality characteristics that are decision-relevant and that have a major influence on service quality as perceived by an end user. In a first step, all quality characteristics are identified from related international standards or specifications ("identification"). Those that are essential and decision-relevant for end user perception are chosen next ("rating"). They have to be selected for evaluation ('selection') and quantified ("quantification") before simple, objective, and cost-effective evaluation methods can be used ("evaluation") in the last step. 2. A Quality of Service evaluation architecture Quality of Service evaluation has to be automated to be successful and to provide usable results. An appropriate architecture involving end users, evaluation meth-ods, and the applications and services under test is designed. The usage of intelligent agents is proposed to do automated Quality of Service evaluation for applications and services. The architecture introduced includes single agent building blocks and communication and coordination models for Quality of Service evaluation agents. 3. A Quality of Service evaluation infrastructure A Quality of Service evaluation infrastructure is needed to establish a trusted environment in which end users can do Quality of Service evaluation. Its first design introduces certification authorities and agent registries as infrastructure prerequisites for standardized and authenticated Quality of Service evaluation agents. This thesis is about Quality of Service evaluation and it addresses quality issues of multimedia applications and distributed services. It fulfills end user requirements. It is based on the author's many years of experience in establishing and running value-added services in distributed environments for different customers