The Deeyook patented sensor is based on wireless interferometry:
The difference in carrier phase between two free-space rays is translated into an angle: the direction of departure.
THE DEEYOOK POSITIONING SYSTEM IS SELF-LEARNING.
Every sensor continuously reviews all location information within its vicinity and locates OFDM APs’ (WIFI and later LTE & 5G ) exact positions (with a 10-cm accuracy rate).
The exact position of each AP is stored in the cloud database.
A global database of APs is created that allows a new way of mapping.
Google has spent billions of dollars wardriving to create a database of APs that features a 5-30-meter accuracy rate. A comparable Deeyook database is self-learning, self-correcting with a 4 inch accuracy rate.
MACHINE LEARNING ADVANCED POSITIONING
Unsupervised estimation of AP parameters using multivariate optimization
Location prediction using Bayesian probability maps based on physical modeling and sensitivity analysis stitching of nearby mapped areas
Embedding in WGS 84 coordinates
EXISTING CELLULAR MODEM TO IMPLEMENT WIFI LOCATIONING
Proprietary firmware can use cellular (LTE or 5G) modem to enable WIFI positioning
ALTERNATIVE DEPLOYMENT: DIRECTION OF ARRIVAL
The Technology is equally effective in reverse (The Deeyook firmware is in the infrastructure OFDM modem). Signal arrives at two antennas from the same source (mobile) and calculation happens at the infrastructure OFDM modem