A Distributed Quality of Service Index Framework

To efficiently support customers to find web service which satisfies the non-functional requirement, a distributed quality of service (QoS) index framework is proposed in this paper. The framework consists of two kinds of structured distributed system, one is chord overlay network which is assigned to classify the web services according to the similarity of web services descriptions, the other is a tree-structured distributed network named DQoS tree which manages the classified web services' QoS. Since there is a dominance relation among multi-dimension QoS, therefore, a guideline based on this relation to construct a QoS tree managed by a sub peer is designed. Through the dominance relation, it's easily find the qualified QoS among web services with less compare times, thus, the query effectiveness is guaranteed. Additionally, to avoid the load imbalance of query load in the DQoS tree, a load balance strategy is introduced which can balance work load efficiently. Experimental results validate the effectiveness and efficiency of the proposed approaches.