Energy-efficient data aggregation techniques for exploiting spatio-temporal correlations in wireless sensor networks

The energy-efficient data aggregation in Wireless Sensor Networks (WSNs) has grown as one of the promising area in many applications. Prior research works have suggested several spatio-temporal models for effectively reducing the data collection costs, but the models are limited to their specificity. This paper presents a pre-filtration method in every sensor node. While using pre-filtration techniques on node's data pre-processing can suppress a huge amount of redundant or correlated data transmissions, thus maximizes the nodes battery lifetime. This paper proposes a framework using both relative variation (RV) and a data aggregative window function (DAWF), which can exploit both spatial and temporal redundancies in WSNs. However, both compressive and prediction-based approaches can find temporal data redundancies or correlations in sensor nodes as well as spatio-temporal correlations in Cluster-Head (CH) nodes. In this regard, the experimental study shows that the proposed mechanism among the considered network nodes can suppress a huge amount of required data transmissions, while providing reliable data towards the base station (BS).