基于策略的动态自适应中间件系统设计

Policy-based dynamic adaptive middleware system design

为满足移动计算的高要求,在参考Ponder策略语言和上下文感知技术的基础上,提出了一种基于策略的动态自适应的中间件系统,以适应移动应用中灵活的、异构的、变化反复无常的上下文执行环境.该系统能根据所获得的上下文采用适当的自适应策略,为移动计算提供一个高度可配置和自适应的执行环境.该系统对策略规则进行定义,并设计基于这些策略的系统自适应性,包括系统的总体结构、自适应策略描述等.该系统已开发实现,并且配置在Java平台上,可完成中间件的交互功能,为以后进一步模拟现实可变环境下的中间件系统提供了可信的实验数据.

To meet the high demand of mobile computing, and adapt to context implementation environmental in heterogeneous and changing frequently mobile environment, a middleware system is designed on the basis of policy and dynamic adaptive technology, which loads Ponder policy language and context-aware technology with references. The system can select appropriate adaptive strategies in accordance with obtained context and provide a highly configurable and adaptive implementation environment for mobile computing. The definition of pohcy rules and its self-adaptive are designed, including the system＇s overall structure, description of adaptive strategies. The system has finished and configured in Java platform, and it can completed interactive functions of middleware, furthermore, system provides reliable experimental data for further simulation of middleware system in variable reality environment.