Forecasting and skipping to Reduce Transmission Energy in WSN

TL;DR

This paper introduces a method to enhance energy efficiency in wireless sensor networks by reducing sensor transmission power through data prediction and request management, leveraging time series forecasting and neural networks.

Contribution

It proposes a novel data reduction and request management approach that minimizes sensor transmission power by utilizing predictions, improving energy efficiency in WSNs.

Findings

Significant reduction in sensor transmission power achieved.

Effective data reduction through prediction-based request management.

Enhanced energy efficiency in WSNs demonstrated.

Abstract

This paper deals with the improvement of energy efficiency in wireless sensor networks (WSN).Taking into consideration the power saving problem which is of crucial importance when the sensors are supplied by a limited power source, this paper proposes a method that optimizes as much as possible the transmission power of the sensors. Under the assumption of perfect channel between the Base Station (BS) and the Sensor Nodes (SN's) and with sufficient power resources at the BS, transferring the effort of transmission power to the BS will not be a peculiar issue. This paper proposes a method that reduces the transmitted data at the SN's and compensate this by requesting the variance of the measured value form predicted values. Furthermore, a request management algorithm (RMA) is developed to reduce the amount of requested data based on consecutive successive predictions. The result of this method reveals a major impact on reducing the transmission power at the SN's. Keywords: Wireless Sensor Networks, Data Reduction, Time Series Forecasting, Artificial Neural Networks.