Physical-MAC Layer Integration: A Cross-Layer Sensing Method for Mobile UHF RFID Robot Reading States Based on...

Area of Science:

• Robotics and Automation
• Wireless Communication and Sensing
• Supply Chain Management

Background:

• Mobile UHF RFID robots in warehouses often follow fixed paths, lacking environmental perception.
• This leads to issues like missed RFID tag reads and inefficient inventory management.

Purpose of the Study:

• To develop a cross-layer sensing method for mobile UHF RFID robots to improve reading accuracy and efficiency.
• To enable robots to perceive shelf layouts and goods distribution for optimized operations.

Main Methods:

• Proposed a physical layer position sensing model using Multiple Linear Regression-Orthogonal Least Squares (MLR-OLS) for shelf localization.
• Developed K-means-based tag density classification and a Random Forest-based missing tag estimation algorithm using physical and MAC layer features.
• Introduced an adaptive reading strategy based on sensed read states to target missing goods.

Main Results:

• Achieved centimeter-level shelf positioning accuracy at medium and low robot speeds.
• Demonstrated high reliability in sensing goods distribution and estimating missing goods counts.
• The adaptive reading strategy significantly improved the overall goods read rate.

Conclusions:

• The proposed cross-layer sensing method enhances the perception and efficiency of mobile UHF RFID robot systems.
• This work provides a foundation for advanced applications of RFID robots in automated warehousing.