A Passive RF Testbed for Human Posture Classification in FM Radio Bands

Area of Science:

• Radio Frequency (RF) Engineering
• Human-Computer Interaction
• Signal Processing

Background:

• Classifying human posture indoors is challenging.
• Existing methods may require dedicated sensors or active signals.
• Leveraging ambient signals like FM radio offers a passive sensing opportunity.

Purpose of the Study:

• To explore the feasibility of classifying human posture using passive FM radio signals.
• To present a novel passive RF testbed for human posture classification experiments.
• To develop and evaluate a methodology for posture detection and classification.

Main Methods:

• Utilizing a passive RF testbed operating in FM radio bands.
• Implementing a methodology involving feature engineering and traditional classification techniques.
• Deploying the system on software-defined radio devices for real-time evaluation.

Main Results:

• Demonstrated the capability of classifying human posture in indoor environments.
• Achieved classification accuracy of approximately 90%.
• Validated the effectiveness of the passive RF testbed for real-time posture analysis.

Conclusions:

• The proposed passive RF testbed effectively classifies human posture using FM radio signals.
• The methodology shows significant potential for developing innovative, passive sensing techniques.
• Future research can build upon this testbed for advanced human activity recognition.