An Environmentally Aware, Intelligently Controlled System for Power Efficient Wireless Sensor Networks

This paper examines the advantages of applying an ANN (artificial neural network) and other machine learning techniques to a battery-powered WSN (wireless sensor network) with the goal of extending network deployment lifetime. Preliminary experimental results have shown that through the use of these techniques that it is possible to achieve network lifetime extensions of several orders-of-magnitude versus always-on systems. (All experimental results were obtained by observing the power consumption of a "live" sensor network deployed using MotelV TelosB sensors in a laboratory environment.).