Granted Patent B2

US 12,520,112 · App. 17/524,024 · Granted Jan 6, 2026

Inventors: Debarshi Patanjali Ghoshal (Montreal, CA); Negar Ghourchian (Montreal, CA); Michel Allegue Martinez (Terrebonne, CA)

Assignee: AERIAL TECHNOLOGIES INC.

H04W4/33H04W4/029

Abstract

Claims (38)

1 . A method for locating a user, the method comprising:

storing data from a plurality of sensing devices in a wireless signal database, each of the plurality of sensing devices located within a sensing area;

polling the wireless signal database for new wireless signal data;

applying a plurality of filters to the new wireless signal data;

storing the filtered wireless signal data in a training data database;

polling a pre-trained proximity database that stores proximity index data;

adjusting the pre-trained proximity database based on shifts detected in the wireless signal data over time;

comparing the new wireless signal data with the proximity index data to determine a current location of a user within the sensing area;

quantifying the proximity index data based on coordinates mapped to the proximity index data;

correlating the quantified proximity index data to movement; and

storing the quantified proximity index data in the pre-trained proximity database.

2 . A system for locating a user, the system comprising:

a plurality devices located within a sensing area;

a wireless signal database that stores raw wireless signal data from the plurality of devices;

a training data database that stores filtered wireless signal data;

a pre-trained proximity database that stores proximity index data,

a wireless communication interface that polls the wireless signal database for new wireless signal data and poll the pre-trained proximity database; and

a processor that executes instructions stored I memory, wherein the processor executes the instructions to:

apply a plurality of filters to the new wireless signal data,

adjust the pre-trained proximity database based on shifts detected in the wireless signal data over time, and

compare the new wireless signal data with the proximity index data to determine a current location of a user within the sensing area; wherein

the plurality of filters remove noise and normalize the wireless signal data;

the plurality of devices are transmitters and receivers;

the training data database stores feature tables from pre-recorded CSI data from a test environment.

3 . The system according to claim 2 , wherein the CSI data was recorded from empty captures and human presence detections in different parts of the test environment.

4 . A method for locating a user, the method comprising:

storing data from a plurality of sensing devices in a wireless signal database, each of the plurality of sensing devices located within a sensing area;

polling the wireless signal database for new wireless signal data;

applying a plurality of filters to the new wireless signal data;

storing the filtered wireless signal data in a training data database;

polling a pre-trained proximity database that stores proximity index data;

adjusting the pre-trained proximity database based on shifts detected in the wireless signal data over time; and

comparing the new wireless signal data with the proximity index data to determine a current location of a user within the sensing area; wherein

the plurality of filters remove noise and normalize the wireless signal data;

the plurality of devices are transmitters and receivers;

the training data database includes feature tables from pre-recorded CSI data from a test environment.

5 . The method according to claim 4 , wherein

the pre-recorded CSI data was recorded for empty captures and human presence captures in different parts of the test environment.

Assignments (2)

Continuity (3)

References Cited (156)