Wireless sensor network abnormal event detecting method based on multi-attribute correlation

Abstract

The invention discloses a wireless sensor network abnormal event detecting method based on multi-attribute correlation and belongs to an anomaly detection technology in data mining.The wireless sensor network abnormal event detecting method comprises the specific steps that non-space-time attribute dependency model is established based on a Bayesian network, the attribute correlation confidence is calculated according to a conditional probability table, the similarity of points to be detected and abnormal points in a non-space-time attribute correlation mode is reflected; time correlation detection is performed based on a sliding window model, readings simultaneously meeting the time correlation mode and abnormal event attribute correlation mode are marked as temporal abnormal points to detect whether abnormal events occur or not by cooperating with neighbor node information.In addition, the wireless sensor network abnormal event detecting method calculates recent abnormal nodes closest to an abnormal event area center by adopting a Ritter's smallest enclosing circle algorithm, uploads the abnormal information of the abnormal nodes and can effectively reduce the data transmission amount and reduce the energy consumption of sensor nodes.The wireless sensor network abnormal event detecting method can be applied to wireless sensor network event detection of multiple sending components.

Description

technical field [0001] The invention relates to the field of computer data mining and application, in particular to a method for detecting abnormal events in a wireless sensor network based on multi-attribute association. Background technique [0002] As one of the basic tasks of data mining, anomaly detection aims to detect data objects that deviate significantly from normal data or normal patterns, so as to reveal rare phenomena and events and discover interesting patterns. In wireless sensor network applications, due to the failure of the node's own hardware and the interference of external environmental noise, the sensor may sense abnormal readings. The wireless sensor network needs to detect the abnormality of the monitoring object in time, and determine the abnormal type, the cause of the induction, the possible consequences and other information, so as to provide decision support for the monitoring personnel. Therefore, how to effectively detect different types of ab...

AI Technical Summary

Problems solved by technology

[0004] The object of the present invention is a method for detecting abnormal events in wireless sensor networks based on multi-attribute association, to solve the problem of high detection false alarm rate caused by ignoring non-spatial-temporal attribute associations in the detection of abnormal events in wireless sensor networks, and to effectively improve the detection rate of abnormal events. Accuracy of event detection

Images