Method and apparatus for reporting channel state information

1. A method for generating channel state information (CSI) reports by a user equipment (UE), the method comprising:

determining a plurality of radio parameters for a connection between the UE and a base station;

computing values of CSI feedback parameters at a current time instance based on the determined plurality of radio parameters;

predicting probable values of the CSI feedback parameters at a future time instance;

generating a CSI report, by compiling at least one of the computed values of the CSI feedback parameters and the predicted values the CSI feedback parameters; and

transmitting the CSI report to the base station,

wherein the UE comprises a neural network, and

wherein the neural network is trained to minimize a cost function, for fitting the generated CSI report between periodic CSI reporting slots.

2. The method of claim 1 , wherein the plurality of radio parameters comprises at least one of at least one CSI-Reference Signal (CSI-RS), or at least one Synchronization Signal Block (SSB).

3. The method of claim 1 ,

wherein the cost function is minimized in case that the predicted values of the feedback parameters at the future time instance match actual values of the feedback parameters at the future time instance, and

wherein the neural network is trained in one of the UE and a cloud server.

4. The method of claim 1 , further comprising:

generating a plurality of CSI reports by at least one of computing or predicting probable values of the feedback parameters at a plurality of future time instances, and,

wherein the plurality of CSI reports are transmitted to the base station in one of a single reporting slot and a plurality of periodic reporting slots.

5. The method of claim 1 , wherein generating the CSI report comprises encoding the compiled at least one of the computed values of the CSI feedback parameters and the predicted values the CSI feedback parameters.

6. The method of claim 1 , wherein computing the values of CSI feedback parameters comprises:

computing reference signal received power (RSRP) associated with a plurality of beams of the UE at the current time instance, for determining an optimal beam at the current time instance and predicting optimal beams at the future time instances.

7. The method of claim 1 , further comprising:

determining an optimal CSI-RS resource allocation and an optimal periodicity for transmitting CSI reports to the base station, based on variation in a condition of a channel between the UE and the base station; and

transmitting, by the UE, to the base station, the optimal CSI-RS resource allocation and the optimal periodicity for transmitting CSI reports to the base station.

8. The method of claim 1 ,

wherein the UE comprises a neural network, and

wherein the neural network model is one of: a hierarchical model, a cascaded model, a joint neural network model, or a deep neural network (DNN) model comprising of fully connected layers and convolutional layers.

9. The method of claim 1 ,

wherein one of the CSI feedback parameters is a precoding matrix indicator (PMI),

wherein a most probable PMI value is determined amongst a predefined number of probable PMI values, and

wherein the predefined number of PMI values are selected amongst a plurality of predicted PMI values.

10. The method of claim 1 ,

wherein one of the CSI feedback parameters is a PMI, and

wherein predicting probable values of the CSI feedback parameters comprises:

determining a dominant Eigen vector amongst a plurality of Eigen vectors, wherein the dominant Eigen vector is determined based on a plurality of feature vectors generated based on content derived from at least one of at least one CSI-RS and at least one SSB;

quantizing the dominant Eigen vector to a nearest valid PMI value, wherein the base station provides a plurality of valid PMI values to the UE; and

generating at least one of a discrete fourier transform (DFT) beam, amplitude coefficients and phase coefficients, based on the quantized dominant Eigen vector, wherein the predicted value of PMI comprises at least one of the DFT beam, amplitude coefficients and phase coefficients.

11. The method of claim 1 , wherein the probable values of the CSI feedback parameters at the future time instance is determined based on a plurality of radio resources received at a plurality of time instances.

12. The method of claim 1 , wherein the plurality of radio parameters are one of measured by the UE and received from the base station.

13. A user equipment (UE) for generating channel state information (CSI) reports, the UE comprising:

a transceiver;

a processor; and

a memory configured to store instructions that, when executed by the processor, cause the processor to:

determine a plurality of radio parameters for a connection between the UE and a base station,

compute values of CSI feedback parameters at a current time instance based on the determined plurality of radio parameters,

predict probable values of the CSI feedback parameters at a future time instance,

generate a CSI report, by compiling at least one of the computed values of the CSI feedback parameters and the predicted values the CSI feedback parameters; and

control the transceiver to transmit the CSI report to the base station,

wherein the UE comprises a neural network, and

wherein the neural network is trained to minimize a cost function, for fitting the generated CSI report between periodic CSI reporting slots.

14. The UE of claim 13 , wherein the plurality of radio parameters comprises at least one of at least one CSI-Reference Signal (CSI-RS), or at least one Synchronization Signal Block (SSB).

15. The UE of claim 13 ,

wherein the processor is further configured to generate a plurality of CSI reports by at least one of computing or predicting probable values of the feedback parameters at a plurality of future time instances, and

wherein the plurality of CSI reports are transmitted to the base station in one of a single reporting slot and a plurality of periodic reporting slots.

16. The UE of claim 13 ,

wherein the cost function is minimized in case that the predicted values of the feedback parameters at the future time instance match actual values of the feedback parameters at the future time instance, and

wherein the neural network is trained in one of the UE and a cloud server.

17. The UE of claim 13 , wherein the processor is further configured to encode the compiled at least one of the computed values of the CSI feedback parameters and the predicted values the CSI feedback parameters.

18. The UE of claim 13 , wherein the processor is further configured to compute reference signal received power (RSRP) associated with a plurality of beams of the UE at the current time instance, for determining an optimal beam at the current time instance and predicting optimal beams at the future time instances.

19. The UE of claim 13 , wherein the processor is further configured to:

determine an optimal CSI-RS resource allocation and an optimal periodicity for transmitting CSI reports to the base station, based on variation in a condition of a channel between the UE and the base station, and

transmit, by the UE, to the base station, the optimal CSI-RS resource allocation and the optimal periodicity for transmitting CSI reports to the base station.

20. The UE of claim 13 ,

wherein the neural network model is one of: a hierarchical model, a cascaded model, a joint neural network model, or a deep neural network (DNN) model comprising of fully connected layers and convolutional layers.