Abstract: A dynamic configuration framework for middleware Quality of Service(QoS) based on reflection is proposed. On basis of dynamic sensing environment change, QoS metric function and Climbing Algorithm(CA) are used to search optimal configuration, and dynamically adjust system parameters to meet users’ service demands by means of reflective mechanism, so that system performance optimization can be achieved online. Simulation experimental results show this framework is effective.

Key words: middleware, reflection, Quality of Service(QoS), Climbing Algorithm(CA)

摘要： 提出一种基于反射的中间件服务质量(QoS)动态配置框架，在动态感知环境变化的基础上，利用QoS度量函数以及爬山算法寻找最佳资源配置，并采用反射机制动态调整中间件系统参数，以满足用户服务需求，从而实现系统在线性能的优化。仿真实验结果表明，该框架是有效的。

关键词: 中间件, 反射, 服务质量, 爬山算法

CLC Number: 

• TP311.5