Conflict Detection in Interval-based Sequences from Wireless Sensor Networks
Pu-Tai Yang, Chih-Jui Chen · 2017
Recently, data mining for wireless sensor networks (WSNs) has received a great amount of research attention due to its wide application potential. Because of the diversity of sensed WSN data, valuable information can be extracted using different data mining techniques, providing many varied insights. This study proposes a framework for analyzing abstracted data obtained by WSNs. In particular, an integrated model for detecting conflicts among sensed interval-based sequences generated by sensors is proposed. Interval-based sequences are results abstracted from sensed data in WSNs. For example, a sensor detects the following interval-based sequence during a time interval: (Al < Bl < Au < Bu), where Al and Bl are the lower limits of events A and B, respectively, whereas Au and Bu are the upper limits. The interval-based sequences can be abstractly expressed and transformed from the sensed data under some pre-defined assumptions. In addition, conflict among two sensed interval-based sequence, a novel notion is proposed The objective of this study is to construct a novel model to detect conflicts and to locate the related sensors within a WSN.