Font Size: a A A

Research And Application Of Frequent Patterns Mining Algorithm In The Wireless Sensor Networks

Posted on:2010-03-06Degree:MasterType:Thesis
Country:ChinaCandidate:S Y MeiFull Text:PDF
GTID:2178360275984511Subject:Software engineering
Abstract/Summary:PDF Full Text Request
In pace with rapid development and increasingly maturation in communication, embedded computing and sensing technology communcation and computation appears in the world. Networks of each devices typically consist of tens or hundreds of small, power constrained nodes which they are expected to monitor for months or years at a time. How to process the continuous large data streams in sensor networks efficiently and how to find interesting knowlede in these streams become new challenge. This thesis mainly studies frequent patterns mining algorithms on applying wireless sensor network.In this thesis, we proposed a novle algorithm based on FP-Stream for this system, which is more suitable for current applications. By the use of Low-frequency Keywords Filtering, filtering unfrequent items in streams, the mining algorithm has been more targeted. By dividing the algorithm into two phases in the mining process, the efficiency of the algorithm is accelerated, but keeping the original precision.In addition, we proposed another frequent pattern mining algorithm for specific application environments in wireless sensor networks, with a compression dictionary tree structureto reduce the memory occupied by the datas; with the use of statistical time window to control the input of data flow; as well as using vector sorting to locate data, so as to improve the algorithm efficiency. With the comparison of experiment, the algorithm perfoms well both in time and space complexity.Finally, we develop a prototype system of frequent pattern mining in wireless sensor networks named SFPMing. It receives the stream of key-word transaction as input, and then mines frequent item set in the data stream and gives immediate resu. Two different architectures for applied circumstance are introducedsed. One is for integrated applications and the distributed one is for enterprise application, Data processing module,monitoring module,storageand modules,logging record module as well as the network transceiver module are illustrated.
Keywords/Search Tags:Frequent Patten, Wireless Sensor Network, Data Mining, Data Stream, Lexicographic Tree
PDF Full Text Request
Related items