Granted Patent B2

US 12,739,009 · App. 18/980,927 · Granted Sep 15, 2026

Inventors: Timur Kochiev (Irvine, CA); Relin Thomas (Seattle, WA)

Assignee: T-Mobile USA, Inc.

H04B7/0695H04W8/18H04W72/56

Abstract

Claims (79)

1 . A non-transitory, computer-readable storage medium comprising instructions recorded thereon, wherein the instructions, when executed by at least one data processor of a system, cause the system to:

receive activity data generated from a wireless device connected to a base station of a telecommunication network,

wherein the activity data includes an account status of a subscriber, subscriber traffic type, or service type, and

wherein the account status of the subscriber is classified as a premium subscriber or a standard subscriber;

cause the activity data to be inputted into a machine learning (ML) model;

receive, as an output from the ML model, a beamforming plan for the base station,

wherein the beamforming plan includes a direction to transmit a beam, a time period to transmit the beam, and at least one network slice on which to transmit the beam; and

generate the beam on the at least one network slice in accordance with the beamforming plan.

2 . The non-transitory, computer-readable storage medium of claim 1 , wherein the wireless device is associated with an account status classified as a premium subscriber, and wherein the system is further caused to:

cause activity data of a second wireless device connected to the base station to be inputted into the ML model,

wherein the second wireless device is associated with an account status classified as a standard subscriber;

receive a second beamforming plan for the second wireless device; and

generate a second beam on a different network slice for the second wireless device,

wherein the beam supplies higher bandwidth speeds or lower latency compared to the second beam on the different network slice.

3 . The non-transitory, computer-readable storage medium of claim 1 , wherein the system is further caused to:

receive an updated beamforming plan based on the wireless device moving to a location outside of a range of the beam; and

implement the updated beamforming plan.

4 . The non-transitory, computer-readable storage medium of claim 1 , wherein the system is further caused to:

receive historical activity data for the wireless device from the base station,

wherein the historical activity data includes historical subscriber traffic type, historical service type, or historical location data of the wireless device; and

train the ML model on the historical activity data.

5 . The non-transitory, computer-readable storage medium of claim 1 , wherein the ML model is located at the base station.

6 . The non-transitory, computer-readable storage medium of claim 1 , wherein the ML model is located on a core node of the telecommunication network.

7 . The non-transitory, computer-readable storage medium of claim 1 , wherein the system is further caused to:

cause the base station to use a zero-forcing beamforming method to generate the beam; and

reduce, using the zero-forcing beamforming method, a side lobe of the beam.

8 . The non-transitory, computer-readable storage medium of claim 7 , wherein the system is further caused to:

reduce, using the zero-forcing beamforming method, a horizontal width of a main lobe of the beam,

wherein reducing the width of the main lobe reduces interference from other wireless devices.

9 . The non-transitory, computer-readable storage medium of claim 1 , wherein the system is further caused to:

prevent the base station from transmitting a wireless signal in all directions on the at least one network slice; and

reduce an amount of greenhouse gases by using a lower amount of resources to direct the beam toward the wireless device.

10 . A system comprising:

at least one hardware processor; and

at least one non-transitory memory storing instructions, which, when executed by the at least one hardware processor, cause the system to:

receive, from a base station of a telecommunication network, activity data generated from a wireless device connected to the base station,

wherein the activity data includes a network traffic type generated by the wireless device;

input the activity data into a machine learning (ML) model;

generate, as an output from the ML model, a beamforming plan for the base station,

wherein the beamforming plan includes a direction to transmit a beam, a time period to transmit the beam, and at least one network slice to transmit the beam on; and

cause the base station to generate at least one beam on a network slice based on the beamforming plan.

11 . The system of claim 10 , further caused to:

receive, from the base station, activity data generated from a second wireless device connected to the base station;

generate, as a second output from the ML model, a second beamforming plan for the base station; and

cause the base station to generate a second beam based on the second beamforming plan on a second network slice.

12 . The system of claim 10 , further caused to:

generate an updated beamforming plan based on the wireless device moving to a new location outside of a range of the beam; and

cause the base station to adjust the direction in which the beam is transmitted based on the new location.

13 . The system of claim 10 , further caused to:

cause the base station to use a zero-forcing beamforming method to generate the beam; and

reduce, using the zero-forcing beamforming method, a side lobe of the beam.

14 . The system of claim 13 , further caused to:

reduce, using the zero-forcing beamforming method, a horizontal width of a main lobe of the beam,

wherein reducing the width of the main lobe reduces interference from other wireless devices.

15 . A method comprising:

receiving real-time activity data generated from a wireless device connected to a base station of a telecommunication network,

wherein the activity data includes an account status of a subscriber, subscriber traffic type, or service type;

causing the activity data to be inputted into a machine learning (ML) model;

receiving, as an output from the ML model, a beamforming plan for the base station,

wherein the beamforming plan includes a direction to transmit a beam, a time period to transmit the beam, and at least one network slice on which to transmit the beam; and

generating the beam on the at least one network slice in accordance with the beamforming plan.

16 . The method of claim 15 , further comprising:

causing activity data of a second wireless device connected to the base station to be inputted into the ML model;

receiving a second beamforming plan for the second wireless device; and

generating a second beam on a different network slice for the second wireless device,

wherein the beam supplies different bandwidth speeds or latency compared to the second beam on the second network slice.

17 . The method of claim 15 , further comprising:

receiving an updated beamforming plan based on the wireless device moving to a location outside of a range of the beam; and

adjusting the direction in which the beam is transmitted based on the updated beamforming plan.

18 . The method of claim 15 , further comprising:

receiving historical activity data for the wireless device from the base station,

wherein the historical activity data includes historical subscriber traffic type, historical service type, or historical location data of the wireless device; and

causing the ML model to be trained on the historical activity data.

19 . The method of claim 15 , further comprising:

causing the base station to use a zero-forcing beamforming method to generate the beam; and

reducing, using the zero-forcing beamforming method, a side lobe of the beam.

20 . The method of claim 19 , further comprising:

reducing, using the zero-forcing beamforming method, a horizontal width of a main lobe of the beam,

wherein reducing the width of the main lobe reduces interference from other wireless devices.

Assignments (1)

Continuity (1)

References Cited (45)