Water level monitoring based on the acoustic signal using the neural network

There is a significant spectrum of Wireless Sensor Networks applications. Using of the WSN fits to natural setting because it does not require supply power sources and communication lines. When proposing such an application the energy management has to be taken into the consideration. This paper is dedicated to the solutions where the energy of the water streams is evaluated based on its acoustic emission in order to predict possibility of floods. For the multi-environment set up the theory of learning systems was inspirational. The ultrasonic sensor works as the reference for learning phase. Evaluation of the acoustic emissions is provided by the neural network. Firstly the system requires the parameter adjustment (learning phase). Afterwards the unit takes into consideration the learned information and apply them in the operational mode. The devices are able to communicate via the RF modules working within the ISM band.