Abstract
With the increasing popularity of utility-oriented computing where the resources are traded as services, efficient management of quality of service (QoS) has become increasingly significant to both service consumers and service providers. In the context of distributed multimedia content adaptation deployment on service-oriented computing, how to ensure the stringent QoS requirements of the content adaptation is a significant and immediate challenge. However, QoS guarantees in the distributed multimedia content adaptation deployment on service-oriented platform context have not been accorded the attention it deserves. In this paper, we address this problem. We formulate the SLA management for distributed multimedia content adaptation deployment on service-oriented computing as an integer programming problem. We propose an SLA management framework that enables the service provider to determine deliverable QoS before settling SLA with potential service consumers to optimize QoS guarantees. We analyzed the performance of the proposed strategy under various conditions in terms of the SLA success rate, rejection rate and impact of the resource data errors on potential violation of the agreed upon SLA. We also compared the proposed SLA management framework with a baseline approach in which the distributed multimedia content adaptation is deployed on a service-oriented platform without SLA consideration. The results of the experiments show that the proposed SLA management framework substantially outperforms the baseline approach confirming that SLA management is a core requirement for the deployment of distributed multimedia content adaptation on service-oriented systems.