Web服务QoS的免疫多信号预测模型研究

Multiple-signal prediction model for QoS of Web services inspired by immune system

合理的服务质量(Quality of Service,QoS)预测是W eb服务承诺的重要依据,在很大程度上关系到服务提供者的信誉。为了有效解决服务质量的预测问题,受到免疫系统多信号机制的启发,提出一种服务提供者主动的实时QoS预测模型。在多Agent仿真平台上的模拟实验和分析表明,该模型能够很好地解决实时QoS预测问题,服务提供者能够根据预测结果动态地调整Web服务的QoS承诺,从而有效提高其信誉。

The rational prediction is the foundation of the promised quality of Web service, which greatly affectsthe reputation of the service provider. To solve the problem of Quality of Service （QoS） prediction affeetively, a real-time QoS prediction model inspired by multiple-signal mechanism in immune system is proposed. The proposed model is studied on the multi-agent simulation platform. The experiments and analysis show that service provider can adjust dynamicaly the promise of QoS according to the prediction results, and then improve his reputation.