Time Series Forecasting Energy-efficient Organization of Wireless Sensor Networks

Area of Science:

• Computer Science
• Electrical Engineering

Background:

• Wireless sensor networks (WSNs) face significant energy constraints, challenging their efficiency.
• Energy efficiency is crucial for the widespread application of WSNs, especially in target tracking.

Purpose of the Study:

• To propose an energy-efficient organization method for WSNs using time series forecasting for target tracking.
• To enhance the energy efficiency of WSNs by optimizing sensing and communication.

Main Methods:

• Formulated WSN organization for target tracking with target, multi-sensor, and energy models.
• Utilized multi-sensor fusion for target localization and historical data for trajectory forecasting.
• Applied Empirical Mode Decomposition (EMD) and Autoregressive Moving Average (ARMA) models for trajectory forecasting.
• Implemented a distributed, probability-based awakening mechanism for sensor nodes.
• Employed Ant Colony Optimization (ACO) for routing to minimize communication energy.

Main Results:

• The combination of EMD and ARMA models accurately estimated target positions.
• The proposed organization method significantly enhanced WSN energy efficiency.
• Both operational and communication energy consumption were minimized.

Conclusions:

• The proposed method effectively improves target tracking accuracy and energy efficiency in WSNs.
• The integration of advanced forecasting techniques and optimized network organization offers a viable solution for energy-constrained WSNs.