Experiments in Data Management for Wireless Sensor Networks

In this paper we present our experimental results in data management for wireless sensor networks when an adapted Fuzzy ART model of neural networks is applied. Our system provides high dimension reduction and transfer savings, sending only classified identification number instead of whole sample. The system implementation is based on MicaZ sensor motes and adapted Fuzzy ART model. It was applied in real-time environments and presents an efficient tool for monitoring and visualization of sensor data. Testing, debugging, data robustness and other implementation issues are also discussed.