Predicting the Energy Efficiency in Wireless Sensor Networks using LSTM and Random Forest Method

Aruna Reddy H., Shivamurthy G., Rajanna M.

Abstract: Over wireless sensor networks, consumption of energy due to needless transmitting data is a serious issue (WSNs). By addressing this issue, any station's lifetime can be extended and system practicality for real world applications can be increased. As a result, for WSN's that use low-powered different sensors, energy-efficient information gathering has become a necessity. Data grouping and forecasting approaches based on symmetrical correlation in sensor information can be utilized in order to downplay the complete utilization of energy consumption of the network for sustained collecting data in these situations. We have integrated a group study of Random Forest (RF) Technique, LSTM technique and Particle Filter (PF) which would be used for a efficient method to evaluate and predict the data required by the Sensor nodes to completely minimize the unnecessary data Transmission. Segmentation and data aggregating to each member nodes are used to effectively make data gathering forecasts in WSNs, primarily to reduce the computational overhead cost associated with constructing the prediction model. Simulation trials, comparison, and performance - based assessment in a variety of scenarios reveal that our approach's forecasting accuracy can exceed traditional ARIMA and Kalman filters with Decision tree models, resulting in improved energy consumption due to fewer packet transmissions.

Keywords: Particle Filter, Random Forest, LSTM, WSN, Clustering

How to Cite?: Aruna Reddy H., Shivamurthy G., Rajanna M., "Predicting the Energy Efficiency in Wireless Sensor Networks using LSTM and Random Forest Method", Volume 13 Issue 1, January 2024, International Journal of Science and Research (IJSR), Pages: 805-811, https://www.ijsr.net/getabstract.php?paperid=SR24105145623, DOI: https://dx.doi.org/10.21275/SR24105145623

Download Citation: APA | MLA | BibTeX | EndNote | RefMan