Abstract:

Since the nodes in wireless sensor networks are usually deployed densely, it would lead to the large redundancy of the collected data. The method of fuzzy clustering is proposed and a fuzzy analogical matrix for clustering is established based on the lattice close degree, and then the redundant nodes are chosen according to the criterion that the number of sensors is as few as possible when the maximum information is obtained. Furthermore, a sleep scheduling mechanism is applied for these redundant nodes and experimental results show that this method can improve the effectiveness and accuracy of the sensor system, and prolong the network life cycle significantly.

Key words: wireless sensor networks; fuzzy clustering; data aggregation; cluster