A Multi-Agent Assistant Dynamic Resource Management Model of Composite Service

Yu Wu,Jie Wu,Jiajun Wang,Shiyong Zhang
DOI: https://doi.org/10.1109/cise.2009.5364507
2009-01-01
Abstract:The paper proposes a multi-agent assistant model (MAaDRMM) to provide a combination QoS management of the composite service in SOA environment. The paper first introduced the QoS model of service resources and its mathematical definitions and then provide MAaDRMM to manage both the relationship between the resources and the runtime status of each resource. When the structure of the work flow is modified or the resource is changed, the model could adjust itself to fit into the new environment, thus providing a dynamic approach to handle the challenges of the management of SOA flexibility. The model also provides an efficient way to manipulate the status of the resources in a heterogeneous environment by taking advantage of web service based management protocol. The goal of the paper is to help service providers to have a more precise control of the real-time QoS and increase the usage level of the resources. affected. In such environment, it is extremely important for service providers to understand both aspects of the problem. The failure of either kind would lead to disastrous consequences of the QoS and user experience of the whole process. The paper proposed a multi-agent assistant model (MAaDRMM) to provide a combination QoS management of the composite service in SOA environment. The model addresses the challenges of both the resource relationship management and the runtime status. The model could adjust itself to fit into the change of the BPEL process, thus providing the dynamic approach to handle the challenges of the flexibility. The paper is organized as follows. Chapter 2 introduces some related work. Chapter 3 describes the resource QoS model and its formal mathematical definition. Chapter 4 describes the multi-agent assistant resource management model (MAaDRMM) and gives the design of 3 main kinds of agents. They are the requirement analysis agent (RAA), the management agent (MA) and the resource agent (RA). Chapter 5 gives an example of the workflow of the model.
What problem does this paper attempt to address?