Adaptive wireless-powered network based on CNN near-field positioning by a dual-band metasurface

Area of Science:

• Wireless communication and networking
• Metamaterials and metasurfaces
• Robotics and intelligent systems

Background:

• The shift towards wireless connectivity necessitates efficient non-contact power transfer solutions for mobile devices.
• Existing wireless power transfer methods often lack precise localization capabilities, limiting applications for moving terminals.
• Battery-free operation is crucial for enabling pervasive sensing and communication in the Internet of Things (IoT).

Purpose of the Study:

• To propose and demonstrate an adaptive wireless-powered network (AWPN) for simultaneous wireless localization and non-contact power supply.
• To enable battery-free sensing, processing, and communication for mobile terminals.
• To achieve high localization accuracy and robust performance in dynamic environments.

Main Methods:

• Utilizing a dual-band metasurface for adaptive beamforming and power transmission.
• Implementing a single-input single-output (SISO) system combining space-time-coding (STC) and convolutional neural network (CNN) for localization.
• Developing a real-time beam alignment mechanism based on precise position information.

Main Results:

• Achieved localization resolution of 3 cm using the proposed STC-CNN approach.
• Demonstrated over 98% classification accuracy for CNN in environmental perception tasks.
• Confirmed the network's ability to tolerate certain environmental changes, ensuring stable power transmission.

Conclusions:

• The developed AWPN effectively integrates wireless localization and contactless power delivery.
• The system enables battery-free operation for mobile terminals, facilitating advanced IoT applications.
• This research paves the way for advancements in intelligent metasurfaces, IoT, and robotic systems.