Data Reduction Methods for Wireless Smart Sensors in Monitoring Water Distribution Systems
M. I. Mohamed, Wenyan Wu, M. Moniri · Procedia Engineering · 2014
In this paper, we introduce smart sensors for monitoring water distribution systems. To improve the quality of monitoring and prolong the lifetime of the sensor node, there is a need to reduce the amount of data transmitted by predicting the future behaviour of the data while detecting and classifying important events locally in the sensor node, transferring only important data. We propose to reduce the amount of data transmitted between the sensor node and end user by using Principle Component Analysis (PCA) and dual prediction models for the monitoring of hydraulic data based on an Autoregressive Moving Average (ARMA) model.