Spatial data estimation in three dimensional distributed wireless sensor networks

Due to deployment of inflated amount of sensor nodes in three dimensional space, observed data are highly correlated among sensor nodes. Since the data are highly correlated, it produces large quantity of redundant data in the network. To reduce data redundancy, we propose a clustering algorithm called Three Dimensional Event based Spatially Correlated Clustering (3D-ESCC) algorithm. Moreover, to extract more accurate data in each distributed cluster of 3D-ESCC algorithm, we propose an Event based Data Estimation (EDE) model in three dimensional space and compare it with other data estimation models. In distributed wireless sensor networks, it may be possible that due to extreme physical condition (e.g heavy rainfall, high temperature and battery discharge) the sensor nodes fails to operate. In such situation, we are able to develop a data prediction model in distributed cluster in case of node failure. Computer simulations and validations are performed to validate 3D-ESCC algorithm and EDE model.