User equipment downlink transmission beam prediction framework with machine learning
Systems, methods, apparatuses, and computer program products for a user equipment downlink transmission beam prediction framework with machine learning are provided. For example, a method can include indicating a beam prediction capability of a user equipment to a network. The method can also include receiving network antenna configuration identification responsive to the indicated beam prediction capability. The method can further include selecting a model for beam prediction based on the network antenna configuration identification. The method can additionally include receiving a configuration of reference signal resources. The method can also include measuring reference signals based on the configuration of reference signal resources. The method can further include performing beam prediction based on the measured references signals using the selected model. The method can additionally include reporting a beam prediction to the network.
1 . An apparatus, comprising:
at least one processor; and
at least one memory storing instructions that, when executed by the at least one processor, cause the apparatus at least to perform
indicating a beam prediction capability of the apparatus to a network;
receiving network antenna configuration identification responsive to the indicated beam prediction capability;
selecting a model for beam prediction based on the network antenna configuration identification;
receiving a configuration of reference signal resources;
measuring reference signals based on the configuration of reference signal resources;
performing beam prediction based on the measured references signals using the selected model; and
reporting a beam prediction to the network.
2 . The apparatus of claim 1 , wherein the configuration of at least some of the reference signal resources comprise beam indexes or one or more angle identifiers information.
3 . The apparatus of claim 1 , wherein the report comprises beam indexes or one or more angle identifiers.
4 . The apparatus of claim 1 , wherein the instructions, when executed by the at least one processor, further cause the apparatus at least to perform
receiving a channel state information configuration for beam prediction, wherein the reference signals are measured in accordance with the channel state information configuration.
5 . The apparatus of claim 1 , wherein the model is trained on a set of data comprising different beams from the measured references signals and different beams from the beam prediction.
6 . The apparatus of claim 1 , wherein the network antenna configuration comprises numbers of rows and columns of antenna elements in a two-dimensional planar antenna, and antenna element distance in azimuth and elevation domain, respectively, of the antenna elements.
7 . The apparatus of claim 1 , wherein the model comprises a machine learning model trained at the apparatus.
8 . The apparatus of claim 1 , wherein the instructions, when executed by the at least one processor, further cause the apparatus at least to perform
when both a prediction resource set and a measurement resource set are configured to the apparatus, identifying resources in the prediction resource set and measurement resource set with beam indexes or one or more angle identifiers, wherein the performing beam prediction based on the beam indexes or one or more angle identifiers.
9 . The apparatus of claim 1 , wherein the configuration of reference signal resources comprises a measurement resource set or a prediction resource set and the measurement resource set.
10 . The apparatus of claim 1 , wherein beam prediction reported to the network comprises at least one of a channel state information reference signal resource indicator for a reference signal resource in a prediction resource set as a predicted best beam; at least one channel state information reference signal resource indicator and a corresponding predicted reference signal received power for a reference signal resource in a prediction resource set as a predicted best beam; or one or more beam indexes or beam angle identifiers.
11 . An apparatus, comprising:
at least one processor; and
at least one memory storing instructions that, when executed by the at least one processor, cause the apparatus at least to perform
receiving an indication of a beam prediction capability of the apparatus;
providing, to the apparatus, a network antenna configuration identification responsive to the indicated beam prediction capability;
providing, to the apparatus, a configuration of reference signal resources;
receiving a beam prediction from the apparatus based on the network antenna configuration identification and configuration of reference signal resources.
12 . The apparatus of claim 11 , wherein the configuration of at least some of the reference signal resources comprise beam indexes or one or more angle identifiers information.
13 . The apparatus of claim 11 , wherein the report comprises beam indexes or one or more angle identifiers.
14 . The apparatus of claim 11 , wherein the instructions, when executed by the at least one processor, further cause the apparatus at least to perform
providing, to the apparatus, a channel state information configuration identification for beam prediction, wherein the reference signals are measured in accordance with the channel state information configuration.
15 . The apparatus of claim 11 , wherein the network antenna configuration comprises downlink transmission beam angle index information.
16 . The apparatus of claim 11 , wherein the network antenna configuration comprises numbers of rows and columns of antenna elements in a two-dimensional planar antenna, and antenna element distance in azimuth and elevation domain, respectively, of the antenna elements.
17 . The apparatus of claim 11 , wherein the configuration of reference signal resources comprises a measurement resource set or a prediction resource set and the measurement resource set.
18 . The apparatus of claim 11 , wherein beam prediction reported by the apparatus comprises at least one of a channel state information reference signal resource indicator for a reference signal resource in a prediction resource set as a predicted best beam; at least one channel state information reference signal resource indicator and a corresponding predicted reference signal received power for a reference signal resource in a prediction resource set as a predicted best beam; or one or more beam indexes or beam angle identifiers.
19 . A method, comprising:
indicating a beam prediction capability of an apparatus to a network;
receiving network antenna configuration identification responsive to the indicated beam prediction capability;
selecting a model for beam prediction based on the network antenna configuration identification;
receiving a configuration of reference signal resources;
measuring reference signals based on the configuration of reference signal resources;
performing beam prediction based on the measured references signals using the selected model; and
reporting a beam prediction to the network.
20 . The method of claim 19 , wherein the configuration of at least some of the reference signal resources comprise beam indexes or one or more angle identifiers information.