基于节点分布密度的传感器网络查询处理机制

Query processing mechanism based on distributed density of wireless sensor networks nodes

为了使传感器网络在进行数据查询时降低能耗和提高网络生命期,引入了一种分布式查询处理机制。这种机制是先将查询分发到网络后再进行优化,这种方法更具有针对性,优化效果也更明显。分簇路由协议与分布式查询有着天然的结合点。每个簇头相当于传统数据库中的一个索引,负责查询的分析、优化和数据融合。簇头根据本区域的节点分布和数据特性可以自主地选择区域内结构而不受其他区域的影响,这样就可以把每个区域看成一个自治系统,而整个传感器网络就是多个自治系统的集合。结果表明:设计查询处理机制时考虑这些因素可以降低能耗和提高网络生命期。

In order to reduce energy consumption and improve the lifetime of wireless sensor networks （WSNs） when processing queries, a distributed query processing mechanism is proposed. It delivers query first and parses and optimizes query at the node where it should reside. This method is more efficient for optimization. Cluster routing protocol is very suitable for this processing system naturally. Every cluster header is equivalent to the index of a traditional database. Its responsibilities include analysis of query, optimization and data fusion. Cluster header decides the structure of the zone according to sensor node distribution and data features. They can differ from other zones. So every zone is an auto-system. The entire network is a aggregation of autonomous systems. Result shows that processing system is energy-efficient and can improve lifetime of network, which is designed with the consideration of these aspects.