Wireless Telemetry: Channel Characterization and Statistical Imputation of Missing Values
Wireless Telemetry: Channel Characterization and Statistical Imputation of Missing Values

Summary/Abstract: An implemented communication wireless telemetric system is presented. It has been built to serve as a measuring station at Thermopiles hot springs. The present work analyzes the method applying several models for adjusting and optimizing the functions of the system. The system also includes a unit that receives and stores data in an appropriate form, ready to be used for statistics. Data of critical physicochemical parameters are continuously measured, processed and transmitted over a wireless link. Path loss is investigated in order to ensure reliable signal reception. In the examined case study of low SNR region with the integrated in the open field sensors, missing data can be a common occurrence. The development of a mathematical technique, based on best prediction, for mitigating the missing values problem is of great significance in establishing a reliable system of wireless telemetry. A novel technique for handling missing values is proposed. An algorithm that evaluates missing values for imputation has been developed, based on a new general linear model analysis for prediction. The full detailed code, is explained and applied on the data that have been captured by the measuring station.

• Details
• Contents