Granted Patent B2

US 12,640,797 · App. 18/729,238 · Granted May 26, 2026

Inventors: Kalle Petteri Kela (Espoo, FI); Teemu Mikael Veijalainen (Espoo, FI)

Assignee: NOKIA TECHNOLOGIES OY

H04B7/0695H04B7/0632H04B7/088

Abstract

Claims (92)

1 . An apparatus, comprising:

at least one processor; and

at least one memory including computer program code;

the at least one memory and the computer program code configured to, with the at least one processor, cause the apparatus at least to:

determine at least one preferred beam for communication between the apparatus and a network node to be used in preference to a beam measured by the apparatus to have a highest quality;

report an indication of said at least one preferred beam to said network node; and

receive pre-learned model data from said network node,

wherein said pre-learned model data comprises at least one of a Q table and neural network weights.

2 . The apparatus of claim 1 , wherein said reporting comprises at least one of:

report said indication of said at least one preferred beam to said network node within a channel state information report;

report an indication of said beam to said network node within a channel state information report; or

report an indication of a candidate set of beams to said network node within a channel state information report.

3 . The apparatus of claim 2 , wherein the apparatus is further caused to establish said candidate set of beams based on measurements of beams received by the apparatus, and wherein said establishing includes at least one of:

beams received by the apparatus measured to exceed a quality threshold amount within said candidate set of beams; or

beams received by the apparatus measured to exceed a channel condition threshold amount within said candidate set of beams.

4 . The apparatus of claim 2 , wherein said determining comprises determining said at least one preferred beam using a machine learning model, and wherein said machine learning model is configured to determine at least one of:

said at least one preferred beam from among said candidate set of beams; and

said at least one preferred beam from among said candidate set of beams by determining which of said candidate set of beams improve a performance metric.

5 . The apparatus of claim 1 , wherein said determining comprises determining at least one of:

said at least one preferred beam between transmission of channel state information reports; or

said at least one preferred beam based on an indication of said at least one preferred beam received from said network node.

6 . The apparatus of claim 1 , wherein said determining comprises at least one of:

measure at least one of a synchronization signal block and a channel state information reference signal to be used by a machine learning model to determine said at least one preferred beam;

determine said at least one preferred beam using a machine learning model for each specified class of user equipment;

determine said at least one preferred beam using a machine learning model for each traffic flow;

determine said at least one preferred beam using a different machine learning model for each specified quality-of-service identifier of user equipment;

determine said at least one preferred beam using a different machine learning model for one or more quality-of-service class identifiers of user equipment; or

determine said at least one preferred beam using a machine learning model for one or more quality-of-service class identifiers of user equipment and configured to fail determine said at least one preferred beam for other quality-of-service class identifiers of user equipment.

7 . The apparatus of claim 1 , wherein the apparatus is further caused to utilize said at least one preferred beam in preference to said beam for communication between the apparatus and said network node, and said utilizing comprises scheduling at least one of:

communication between the apparatus and said network node using said at least one preferred beam; or

communication between the apparatus and said network node using said at least one preferred beam at least until a next channel state information reporting period.

8 . An apparatus, comprising:

at least one processor; and

at least one memory including computer program code;

the at least one memory and the computer program code configured to, with the at least one processor, cause the apparatus at least to:

determine at least one preferred beam for communication between a user equipment and said apparatus to be used in preference to a beam measured by said user equipment to have a highest quality; and

utilize said at least one preferred beam in preference to said beam for communication between said user equipment and the apparatus,

wherein said determining comprises at least one of:

receive an indication of said at least one preferred beam within a channel state information report from said user equipment;

receive an indication of said beam within a channel state information report from said user equipment;

receive an indication of a candidate set of beams within a channel state information report from said user equipment;

establish said candidate set of beams based on measurements of beams received by the apparatus;

include beams received by the apparatus measured to exceed a quality threshold amount within said candidate set of beams;

include beams received by the apparatus measured to exceed a channel condition threshold amount within said candidate set of beams;

determine said at least one preferred beam between transmission of channel state information reports;

determine said at least one preferred beam based on an indication of said at least one preferred beam received from said user equipment; or

determine said at least one preferred beam using a machine learning model.

9 . The apparatus of claim 8 , wherein said machine learning model is configured to perform at least one of:

determine said at least one preferred beam from among said candidate set of beams;

determine said at least one preferred beam from among said candidate set of beams by determining which of said candidate set of beams improve a performance metric;

determine said at least one preferred beam for all radio resource control connected user equipment with a cell;

determine said at least one preferred beam using a machine learning model for each specified class of user equipment;

determine said at least one preferred beam using a machine learning model for each traffic flow;

determine said at least one preferred beam using a different machine learning model for each specified quality-of-service identifier of user equipment;

determine said at least one preferred beam using a different machine learning model for one or more quality-of-service class identifiers of user equipment; or

determine said at least one preferred beam using a machine learning model for one or more quality-of-service class identifiers of user equipment and configured to fail determine said at least one preferred beam for other quality-of-service class identifiers of user equipment.

10 . The apparatus of claim 9 , wherein said determining of the at least one preferred beam from among said candidate set of beams is based on at least one of:

a probability of scheduling that bearer due to quality-of-service priorities of other user equipment within that cell and their serving beams;

interference by neighboring network nodes and user equipment; and

combined performance of uplink and downlink with that bearer;

determine said at least one preferred beam from among said candidate set of beams based on measured uplink transmission from user equipment;

select and sort user equipment based on a time domain scheduling metric;

determine said at least one preferred beam from among said candidate set of beams for highest priority user equipment;

remove user equipment having different beams;

determine said at least one preferred beam from among said candidate set of beams based on measured uplink transmission from user equipment; or

measure at least one of a synchronization signal block and a channel state information reference signal to be used by the machine learning model to determine said at least one preferred beam.

11 . The apparatus of claim 8 , wherein said utilizing comprises at least one of:

schedule communication between said user equipment and the apparatus using said at least one preferred beam; and

schedule communication between said user equipment and the apparatus using said at least one preferred beam at least until a next channel state information reporting period.

12 . A method, comprising:

determining, by a user equipment, at least one preferred beam for communication between said user equipment and a network node to be used in preference to a beam measured by said user equipment to have a highest quality;

reporting an indication of said at least one preferred beam to said network node; and

utilizing said at least one preferred beam in preference to said beam for communication between said user equipment and said network node, and said utilizing comprises scheduling at least one of:

communication between said user equipment and said network node using said at least one preferred beam; or

communication between said user equipment and said network node using said at least one preferred beam at least until a next channel state information reporting period.

13 . The method of claim 12 , wherein said reporting comprises at least one of:

report said indication of said at least one preferred beam to said network node within a channel state information report;

report an indication of said beam to said network node within a channel state information report; or

report an indication of a candidate set of beams to said network node within a channel state information report.

14 . The method of claim 13 , wherein said determining comprises determining said at least one preferred beam using a machine learning model, and wherein said machine learning model is configured to determine at least one of:

said at least one preferred beam from among said candidate set of beams; and

said at least one preferred beam from among said candidate set of beams by determining which of said candidate set of beams improve a performance metric.

15 . The method of claim 12 , wherein said determining comprises determining at least one of:

said at least one preferred beam between transmission of channel state information reports; or

said at least one preferred beam based on an indication of said at least one preferred beam received from said network node.

16 . The method of claim 12 , wherein said determining comprises at least one of:

measure at least one of a synchronization signal block and a channel state information reference signal to be used by a machine learning model to determine said at least one preferred beam;

determine said at least one preferred beam using a machine learning model for each specified class of user equipment;

determine said at least one preferred beam using a machine learning model for each traffic flow;

determine said at least one preferred beam using a different machine learning model for each specified quality-of-service identifier of user equipment;

determine said at least one preferred beam using a different machine learning model for one or more quality-of-service class identifiers of user equipment; or

determine said at least one preferred beam using a machine learning model for one or more quality-of-service class identifiers of user equipment and configured to fail determine said at least one preferred beam for other quality-of-service class identifiers of user equipment.

Assignments (2)

Priority Claims (1)

Continuity (1)

References Cited (18)