Intelligent data aggregation in sensor networks using artificial neural-networks algorithms

Some of the algorithms developed within the artificial neural-networks tradition can be easily adopted to wireless sensor network platforms and will meet the requirements for sensor networks like: simple parallel distributed computation, distributed storage, data robustness and autoclassification of sensor readings. As a result of the dimensionality reduction obtained simply from the outputs of the neural-networks clustering algorithms, lower communication costs and energy savings can also be obtained. In the paper we will propose three different kinds of architectures for incorporating the ART and FuzzyART artificial neural networks into the small Smart-It units’ network. We will also give some results of the classifications of real-world data obtained with a sensor network of 5 Smart-It units, each equipped with 6 different types of sensors. We will also give results from the simulations where we have purposefully made one of the input sensors malfunctioning, giving zero or random signal, in order to show the data robustness of our approach.