Reducing the data transmission in WSNs using time series prediction model

TLDR

Wireless Sensor Networks continuously monitor physical phenomena but are constrained by limited battery power, prompting research into energy‑saving data reduction schemes. This study implements a data‑reduction scheme based on time‑series prediction to reduce transmissions in WSNs. The scheme employs the GM(1,1) grey prediction model, known for its computational efficiency and accuracy, to forecast sensor data and suppress unnecessary transmissions. Simulations show that achieving ±5% accuracy requires only 43% of transmissions, thereby extending the network lifetime.

Abstract

Wireless Sensor Networks (WSNs) are used for continuously monitoring some physical phenomenon like temperature, humidity etc over a large geographical area. But the limited power supply is the major constraint of the Wireless Sensor Network because it uses non-rechargeable batteries in sensor nodes when it is being used in the areas where human approach is nearly impossible. A lot of researches are going on all over the world to reduce the energy consumption in sensor nodes. Data reduction scheme of energy conservation can be used to reduce the power consumption in WSNs. In this paper, data reduction scheme has been implemented using the prediction based approach. GM(1,1) model is used as the prediction model which is being used worldwide in most of the applications for predicting future values of time series data using some past values due to its high computational efficiency and accuracy. In simulations, it has been seen that for ±5% accuracy, only 43% transmissions are needed, thus increasing the overall lifetime of WSNs.