A Novel Data Prediction Technique Based on Correlation for Data Reduction in Sensor Networks

Abstract

Environmental monitoring is among the most significant applications of wireless sensor networks (WSNs), which results in sensing, communicating, aggregating and transmitting large volumes of data over a very short period. Thus, a lot of energy is consumed in transmitting this redundant and correlated data to the basestation (BS) making it enormously challenging to achieve an acceptable network lifetime, which has become a bottleneck in scaling such applications. In order to proficiently deal with the energy utilization in successive data aggregation cycles, we propose a data prediction-based aggregation model, which will reduce data transmission by establishing relationship between sensor readings. The purpose of the proposed model is to exempt the sensor nodes (SN) from sending huge volumes of data for a specific duration during which the BS will predict the future data values and thus minimize the energy utilization of WSN. The study suggested an extended linear regression model, which determines resemblance in shape of data curve of contiguous data periods. We have used real sensor dataset of 54 SN that was deployed in the Intel Berkeley Research laboratory. We tested and compared our work with the recent prediction-based data reduction method. Results reveal that the proposed ELR model works better when compared with the other techniques in many assessment indicators.