Machine Learning Assisted Position Detection of UHF RFID Tags

Abstract:
A typical problem of UHF RFID installations is the lack of location information. UHF systems are designed for large bulk read processes as opposed to sequential reads typical for optical systems (e.g., Barcode) or short range low frequency systems such as NFC. [1] In optical systems the position of an object can easily be determined because the scanner must be have a light of sight to the targeted tag. However, such information is not available in UHF-based RFID technology. This paper describes a novel method – based on radar technology and machine learning algorithms – that can estimate position of RFID tags using the UHF signals reflected from the tags. Another advantage of this method is the elimination overreaching problems.