# US 2022/0035972 A1 · Systems and methods for denoising gps signals using simulated models

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2022/0035972 A1
- Document ID: US20220035972A1
- Status: Published application
- Kind: A1
- Application number: 16/941,076
- Filed: 2020-07-28
- Published: 2022-02-03
- Earliest claimed: 2020-07-28
- Assignees: Verizon Patent and Licensing Inc.
- Applicants: Verizon Patent and Licensing Inc.
- Inventors: Jerry GAMBLE, JR.
- Web page: https://patentgrep.com/patent/US20220035972A1

## CPC

- G01S 19/07: providing data for correcting measured positioning data, e.g. DGPS [differential GPS] or ionosphere corrections
- G06F 30/27: using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
- G01S 19/23: Testing, monitoring, correcting or calibrating of receiver elements
- G01S 19/393: {Trajectory determination or predictive tracking, e.g. Kalman filtering}
- G01S 19/40: Correcting position, velocity or attitude
- G06N 3/02: Neural networks
- G06N 3/044: Recurrent networks, e.g. Hopfield networks
- G06N 3/0442: characterised by memory or gating, e.g. long short-term memory [LSTM] or gated recurrent units [GRU]
- G06N 20/00: Machine learning
- G01S: RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS

## Abstract

A device may receive simulation information associated with iteratively simulating movement of a geolocation unit. The device may generate simulated true paths for simulations of the movement of the geolocation unit. The device may configure, based on the simulation information, noise parameters associated with simulating distortion of a signal received by the geolocation unit during the simulations to generate simulated measured paths. The device may execute the simulations to correspondingly generate sets of simulation data, training, using the sets of simulation data, a geolocation correction model based on differences between true location data of the simulated true paths and simulated measured data of the simulated measured paths. The device may configure the geolocation correction model to process a geolocation signal of an active geolocation unit to determine a location of the active geolocation unit based on the differences between the true location data and the simulated measured data.

## Claims

### Claim 1

A method, comprising:
receiving, by a device, simulation information associated with iteratively simulating movement of a geolocation unit;
generating, by the device, a plurality of simulated true paths for simulations of the movement of the geolocation unit;
executing, by the device, the simulations to correspondingly generate sets of simulation data;
modifying, by the device, the sets of simulation data to generate modified sets of simulation data representing distortion of a signal received by the geolocation unit during the simulations,
wherein the modified sets of simulation data correspond to simulated measured paths;
training, by the device and using the modified sets of simulation data, a geolocation correction model based on differences between true location data of the plurality of simulated true paths and simulated measured data of the simulated measured paths; and
generating, by the device and based on training the geolocation correction model, a global positioning system (GPS) de-noiser to process a geolocation signal of an active geolocation unit to determine a location of the active geolocation unit based on the differences between the true location data and the simulated measured data.

### Claim 2 (depends on 1)

The method of claim 1, wherein modifying the sets of simulation data, for each simulation of the simulations, comprises:
applying a first level of Gaussian noise to a true simulation path of the simulation for a first time period of the simulation; and
applying a second level of Gaussian noise to the true simulation path of the simulation for a second time period of the simulation,
wherein a maximum of a range of the first level of Gaussian noise is less than or equal to a threshold percentage of a minimum of a range of the second level of Gaussian noise.

### Claim 3 (depends on 2)

The method of claim 2, wherein the first time period is associated with a first set of timestamps of the simulation and the second time period is associated with a second set of timestamps of the simulation,
wherein a quantity of the first set of timestamps is greater than or equal to a threshold percentage of a total quantity of timestamps associated with a combination of the first time period and the second time period, and
wherein a quantity of the second set of timestamps is less than the threshold percentage of the total quantity of timestamps.

### Claim 4 (depends on 1)

The method of claim 1, wherein the modified sets of simulation data indicate the differences based on mapping, based on a simulated timing of an associated simulation, true simulation locations of the true location data to corresponding measured locations of the simulated measured data.

### Claim 5 (depends on 1)

The method of claim 1, wherein modifying the sets of simulation data includes:
utilizing a first filter to modify a first portion of the sets of simulation data; and
utilizing a second filter to modify a second portion of the sets of simulation data,
wherein the first filter is different from the second filter.

### Claim 6 (depends on 1)

The method of claim 1, wherein generating the GPS de-noiser comprises:
enabling the GPS de-noiser to receive, from the active geolocation unit, measured geolocation data associated with the geolocation signal;
processing, using the GPS de-noiser, the measured geolocation data to generate corrected geolocation data;
perform a geolocation analysis of the active geolocation unit based on the corrected geolocation data,
wherein the geolocation analysis is relative to the active geolocation unit being within a threshold distance of a particular geographical location; and
perform one or more actions associated with the active geolocation unit being within the threshold distance of the particular geographical location.

### Claim 7 (depends on 6)

The method of claim 6, wherein performing the one or more actions comprises at least one of:
providing, to a user device associated with the active geolocation unit, a notification that the user device is or was within the threshold distance of the particular geographical location; or
causing an interface of a geolocation platform to indicate that the user device is or was within the threshold distance of the particular geographical location.

### Claim 8

A device, comprising:
one or more processors configured to:
receive simulation information associated with simulating movement of a geolocation unit;
configure, based on the simulation information, a simulation engine to generate a simulation to simulate the movement of the geolocation unit;
train a geolocation correction model associated with correcting a measured geographic location of the geolocation unit by iteratively:
causing the simulation engine to execute an iteration of the simulation based on a set of simulation parameters to generate simulation data,
sorting the simulation data into sets of simulation data that are associated with true simulation locations and corresponding measured simulation locations,
modifying the sets of simulation data to generate training data,
wherein the sets of simulation data are modified to represent noise included in a signal received by the geolocation unit,
training the geolocation correction model based on the training data,
determining a model score associated with the geolocation correction model being trained based on the iteration, and
unless the model score satisfies a convergence threshold, updating the simulation parameters for a subsequent iteration of training the geolocation correction model using a subsequent iteration;
determine, after an iteration of training the geolocation correction model, that the model score satisfies the convergence threshold; and
utilize the geolocation correction model to process a geolocation signal of an active geolocation unit to determine a location of the active geolocation unit.

### Claim 9 (depends on 8)

The device of claim 8, wherein the simulation information identifies a type of the geolocation unit,
wherein simulated true paths, associated with the true simulation locations, and time lengths of iterations of the simulation are generated based on the type of the geolocation unit.

### Claim 10 (depends on 8)

The device of claim 8, wherein the simulation information identifies a geographical area,
wherein simulated true paths, associated with the true simulation locations, are generated based on geographical features of the geographical area.

### Claim 11 (depends on 8)

The device of claim 8, wherein the simulation information identifies a desired level of accuracy associated with the geolocation correction model determining a true location of the active geolocation unit,
wherein the convergence threshold is based on the desired level of accuracy.

### Claim 12 (depends on 8)

The device of claim 8, wherein configuring the simulation engine comprises:
configuring ranges for one or more simulation parameters of the set of simulation parameters based on the simulation information.

### Claim 13 (depends on 8)

The device of claim 8, wherein the set of simulation parameters comprises at least one of:
a first level of Gaussian noise to be applied to simulate distortion of a geolocation signal received by the geolocation unit during the simulation;
a second level of Gaussian noise to be applied to simulate an anomaly in the geolocation signal;
a frequency of applying the second level of Gaussian noise to the geolocation signal;
a distance of a simulation path corresponding to the movement of the geolocation unit; or
a time length of the simulation.

### Claim 14 (depends on 8)

The device of claim 8, wherein the one or more processors, when configuring the geolocation correction model, are configured to at least one of:
deploy the geolocation correction model to the active geolocation unit to permit the active geolocation unit to locally process the geolocation signal; or
enable the geolocation correction model to receive, from the active geolocation unit, measured geolocation data associated with the geolocation signal and process the measured geolocation data to indicate the location of the active geolocation unit.

### Claim 15

A non-transitory computer-readable medium storing instructions, the instructions comprising:
one or more instructions that, when executed by one or more processors, cause the one or more processors to:
receive a geolocation signal associated with a user device;
determine, based on the geolocation signal, a measured location of the user device;
generate measured geolocation data that corresponds to the measured location;
process, using a geolocation correction model, the measured geolocation data to generate corrected geolocation data,
wherein the geolocation correction model is trained based on simulation data associated with a plurality of executed simulations of movement of a geolocation unit associated with the user device; and
perform an action associated with the corrected geolocation data and the user device.

### Claim 16 (depends on 15)

The non-transitory computer-readable medium of claim 15, wherein the simulation data comprises sets of true location data and corresponding simulated measured data of the plurality of executed simulations,
wherein the true location data are associated with simulation paths of the plurality of executed simulations and the corresponding simulated measured data are generated based on applying Gaussian noise to the simulation paths.

### Claim 17 (depends on 15)

The non-transitory computer-readable medium of claim 15, wherein the plurality of executed simulations were executed based on simulation parameters that were configured based on at least one of:
a characteristic of the user device, or
a geographical region that includes the measured location.

### Claim 18 (depends on 15)

The non-transitory computer-readable medium of claim 15,
wherein the geolocation correction model is associated with a geolocation application installed on the user device.

### Claim 19 (depends on 15)

The non-transitory computer-readable medium of claim 15, wherein the geolocation correction model comprises a neural network that was trained based on:
an autoencoder model iteratively configuring simulation parameters of the plurality of executed simulations, or
a long short-term memory model iteratively identifying dependencies associated with simulated measured data and true location data of the plurality of executed simulations.

### Claim 20 (depends on 15)

The non-transitory computer-readable medium of claim 15,
wherein the one or more instructions, that cause the one or more processors to perform the action, cause the one or more processors to at least one of:
indicate, via a display of the user device, a location corresponding to the corrected geolocation data; or
provide, to a geolocation platform, the corrected geolocation data to permit the geolocation platform to perform a geolocation analysis associated with the user device.
