Vector Time Series Identification Methodology Based on Wireless Sensor Networks

摘要

Wireless sensor networks being an assembly of numerous sensor nodes that collect scalar (temperature, humidity, etc.) or vector measurements (velocity, acceleration, force or torque) provide huge opportunities for monitoring and mathematical modeling of the time-evolution for physical quantities under investigation. Base on the localized vector measurements provided by sensor nodes, this paper offers an efficient methodology to identify vector time series in any location within the wireless sensor network coverage. The methodology is confirmed by a simulation case study regarding wind speed prediction.