【作者】 张犁； 孙洁； 潘纲； 吴朝晖；

【Author】 Zhang Li, Sun Jie, Pan Gang+, Wu Zhao-Hui (Department of Computer Science, Zhejiang University, Hangzhou 310027, China)

【机构】 浙江大学计算机学院；

【摘要】 本文给出了一个情境感知的数据管理中间件框架,其包括情境信息的处理机制和根据特定情境获取数据服务的原理。基于此框架,我们采用Java语言及类库,实现了一个情境感知的数据管理平台,并通过实验,总结了开发普适计算环境下系统的一些经验。更多还原

【Abstract】 With the progress of pervasive computing technology, smart space such as smart home, smart classroom and smart workshop are coming into our everyday life and will profoundly change the way we live. Context-awareness is one of the key technologies to support these smart scenarios, which mainly deals with sensing changes in the smart space, and timely activates proper actions to adapt to the current contexts. Context-aware data management is considered as one of the most important and necessary components in a smart space, whose computing capability much depends on its prompt access to appropriate data. This paper focuses on the principles, models, and mechanisms of data management in pervasive computing environment. We present a middleware framework for context-aware data management, and implement a prototype. In our prototype, we roughly divide the concept “data” into two categories: files and structured data (i.e. records in databases). We adopt metadata written in XML to associate context information with these data. In daily life, we often store the files in distributed hosts, e.g. parts of files for a special meeting left at home, others at office. For context-based file access, we can build a metadata file as the directories. If we want to query files in specific cases, the current contexts can be matched with the metadata file to get the handles of the proper files. While querying data records in a database, we develop an algorithm that will translate the current contexts into conditions or constrains, then combine them with the query target and automatically generate the SQL query codes. With this approach, we form the foundation of context-aware accessing of the target data according to the current context. Our middleware provides a infrastructure for context gathering, publishing, and querying, which is an extension of Gatech’s Context-toolkit. We also emphasize scalability and self-adaption in the prototype. For scalability, a developer can easily develop context widgets, servers, interpreters and data services for his own purpose, as well as adding components from other developers into our framework. For self-adaption, we have defined different work modes, e.g. meeting mode and driving mode. The system detects the current work mode, and will dynamically load different data services for adaptation. Besides, we developed an adaptive user interface for users to check the current available services, and to download the services as plug-ins on users’ need.更多还原