Font Size: a A A

Research On Data Aggregation Query Strategy In Wireless Sensor Network

Posted on:2010-12-29Degree:MasterType:Thesis
Country:ChinaCandidate:S ShenFull Text:PDF
GTID:2178360275482521Subject:Information and Communication Engineering
Abstract/Summary:PDF Full Text Request
Wireless sensor network (WSN) is a special network that composed with numerous of sensor nodes which limited with computing, communicating and storage ability. WSN has been the targets of active research in the recent due to their military and civil applications. Nowadays one of the most important utilization in sensor networks is data query. The data query takes the WSN as a large-scale and distributed database, and the query results can be collected effectively which combined with a data-centric routing protocol.Since the way of sensing and collecting data were not assigned in query plan in existing database models of WSN. We can explore multiple implementations from the bottom level in data query processing. This paper was aimed at the optimization strategy of aggregation queries which have widespread in WSN. The main works are as follows:Firstly this paper proposed a routing mechanism that is source-nodes-oriented and suitable to data query. This mechanism includes a query plan propagation algorithm, route maintenance and a copy-mechanism. The query plan propagation algorithm restrains the unnecessary broadcast messages and reserves an initial aggregation tree for each source node. It lays the foundation for the data aggregation in the next stage of data query and it is also the basis of the long-term mechanism which used in the end of life of the query. The goal of the route maintenance mechanism is to maintain the stability in topology. As node maintains more than one parent in the network, we proposed a copy-mechanism to process date packet copies which targets on system robustness, we discussed how to control the copies in the network and how to deal with the copies that arrive late in base station node.Secondly, analyzed the performance of aggregation tree constructing algorithms in three aspects and proposed a global energy-efficient aggregation tree constructing algorithm based on greedy incremental tree. This algorithm got the best candidate's information with few packets exchanged in presence tree and avoided mass exploratory messages broadcasted by sources periodically. The global energy consumption of constructing the aggregation tree was affected by node density slightly and maintained at a relatively low level. Since the construction time was no longer restricted by the exploratory messages'periods, the algorithm can improve the speed of constructing a greedy incremental tree (GIT).Finally, this thesis discussed the long-term mechanism of prolonging query lifetime in two aspects. On one side is the self-adaptive mechanism applies to the situation that the query plan has not appoint the data collection frequency, we discussed how to adjust the data sensing and data sending frequency according to the node's resources. We proposed an interpolation algorithm to resolve the problem that there are not enough data in a data sending cycle. On the other hand, if the collection frequency has been given in the plan and can not be changed, a long-term mechanism that rotating the root node of an aggregation tree among source nodes is also proposed to guarantee the network for maximum vitality.
Keywords/Search Tags:Wireless Sensor Network, Data query, Query optimization strategy, Routing, Data aggregation, Long-term mechanism
PDF Full Text Request
Related items