A Spatio-Temporally Adaptive Clustering Protocol for wireless microsensor networks

We present a spatio-temporally adaptive clustering (STACK) protocol for wireless microsensor networks. It takes into account data correlation, both spatial and temporal, using dynamically computed nearness and activity functions. The objective is to uniformly budget energy expenditure over space and time. All three forms of energy expenditure, namely transmission, reception and computation are considered. The STACK protocol also enhances the lifetime of the network by predicting the survivability of the formed clusters. STACK provides a higher variance for the set of data collected, thus leading to a more accurate picture of the region of interest. Our simulation demonstrates this and achieves better energy efficiency and lifetime compared to existing approaches.