Sensor Data Fusion and Event Detection

Abstract

Sensor data often contains noise, outliers, missing values, and a significant number of duplicate values. The causes of such data quality problems include the sensors’ internal errors, a harsh environment in which the sensors are deployed, and data loss occurring during wireless transmission. Sensor data fusion consists of three steps, data pre-processing, data mining, and data post-processing. This chapter discusses data pre-processing and data mining. Data pre-processing includes data cleaning, outlier detection, missing values recovery, data reduction, and data prediction, etc. Neighbourhood support and tempo-spatial pattern extraction are introduced and applied to a generic sensor state model for event detection. The concept of in-network database is also introduced by presenting WSNs as a virtual distributed database.