Experimental evaluation of data aggregation methods applied to soil moisture measurements

The implementation of precision irrigation systems represent an interesting opportunity in order to increase water efficiency for agricultural, gardens and parks areas. Precision irrigation may include a large group of sensors, one controller unit and few or only one actuator elements. Soil moisture is primarily used as the main process variable in the control loop, meanwhile wireless sensor networks are often required to communicate soil moisture data from different sensor nodes distributed over a usually large irrigation area to the central controller unit. This implies that many and different soil moisture values are simultaneously acquired during the sensing process. However a single representative and consistent soil moisture value for the complete irrigation area is required for control purposes, which can be obtained through a data aggregation process. This paper presents an experimental evaluation of different data aggregation methods applied to soil moisture measurements in a closed-loop precision irrigation system installed in an experimental field.