Abstract: With the large-scale application of Web, the number of distributed data stream is increasing rapidly and the query processing is facing the great challenge. This paper proposes the P2P-based middleware prototype of distributed data stream for real-time answer. By utilizing the content-based routing performance such as scalability, load balance of communication and dynamic adaptability, the solution can handle the query of product and similarity efficiently. Simulation experiment verifies that the indexing mechanism can reduce the computing resources of network link and greatly improve the efficiency of data stream processing.

Key words: data stream, synopsis structure, middleware

摘要： 随着Web的大规模应用，分布式数据流的数量迅速增长，其查询处理面临极大的挑战。该文开发了分布式数据流响应查询的P2P中间件原型，利用基于内容路由所提供的可扩展性、通信负载平衡及动态适应性等特性，能有效地处理相似查询，支持内积查询。模拟实验表明该索引机制能减少网络连接的计算资源，提高数据流查询处理效率。

关键词: 数据流, 概要结构, 中间件

CLC Number: 

• TP311