An Online Charging Scheme for Wireless Rechargeable Sensor Networks Based on a Radical Basis Function.

Area of Science:

• Wireless Sensor Networks
• Energy Management
• Machine Learning

Background:

• Current online charging schemes for wireless rechargeable sensor networks lack dynamic energy consumption rate estimation.
• This deficiency leads to poor charging response and the frequent occurrence of energy holes in nodes.

Purpose of the Study:

• To propose an energy hole avoidance online charging scheme (EHAOCS) using a radical basis function (RBF) neural network.
• To improve the efficiency and reliability of wireless rechargeable sensor networks by mitigating energy holes.

Main Methods:

• Developed an RBF neural network-based online charging scheme (RBF-EHAOCS).
• The RBF network dynamically predicts node energy consumption rates.
• Optimal charging request thresholds are estimated, and charging nodes are selected based on minimal energy hole rate and shortest charging latency.

Main Results:

• The RBF-EHAOCS demonstrated a significantly lower node energy hole rate compared to existing schemes.
• The proposed scheme also achieved shorter charging latency for nodes requiring charging.
• Simulation results validate the effectiveness of the RBF-EHAOCS.

Conclusions:

• The RBF-EHAOCS effectively addresses the energy hole problem in wireless rechargeable sensor networks.
• Dynamic energy consumption prediction using RBF neural networks enhances charging efficiency.
• The proposed scheme offers a superior alternative to existing online charging strategies.