A clustering approximation mechanism based on data spatial correlation in wireless sensor networks

In wireless sensor networks (WSNs), the sensor nodes that locate near often sense the similar data, however, transmitting the repeated or redundant data often cause unnecessary energy consumption. Aiming at this point, this paper firstly proposes a gridbased spatial correlation clustering (GSCC) method which clusters the sensor nodes according to data correlation. According to GSCC, in the same cluster the member nodes have high similarity. Based on GSCC, then this paper proposes a spatial correlation clustering approximation framework (SCCAF). SCCAF can largely save networks´ energy by which the cluster head estimates the data of its member nodes provided that approximation value is in the allowable error range. Experiments prove that not only SCCAF based on GSCC method can prolong the lifetime of the sensor networks compared with LEACH but also SCCAF guarantees more accuracy than CASA (clustering-based approximate scheme for data aggregation) which is a previous approximation scheme.