Neural network-based spatial inter-cell interference learning
A method of wireless communication by a network device includes receiving a location of a user equipment (UE). The method also includes receiving information associated with neighbor cell transmit beams. The method further includes predicting a spatial inter-cell interference-based distribution at the location of the UE based on the information associated with the neighbor cell transmit beams.
1. A method of wireless communication by a network device, comprising:
receiving a location of a user equipment (UE);
receiving information associated with neighbor cell transmit beams; and
predicting a spatial inter-cell interference-based distribution at the location of the UE based on the information associated with the neighbor cell transmit beams.
2. The method of claim 1 , further comprising coordinating with the at least one neighbor cell to schedule downlink transmit beams in time and frequency resources different than time and frequency resources serving the UE.
3. The method of claim 1 , further comprising training a neural network to learn the spatial inter-cell interference-based distribution for a plurality of locations.
4. The method of claim 3 , in which the network device comprises a serving cell and learning the spatial inter-cell interference-based distribution is based on receiving channel state information (CSI) reference signal (CSI-RS) measurement reports from UEs at the plurality of locations.
5. The method of claim 3 , in which the network device comprises a neighbor cell or a centralized node to learn the spatial inter-cell interference-based distribution is based on information received from a serving cell.
6. The method of claim 3 , in which the network device comprises a serving cell and the method further comprises training the neural network at the serving cell to compute neural network weights.
7. The method of claim 6 , further comprising sharing the neural network weights with a neighbor cell.
8. The method of claim 6 , further comprising sharing the neural network weights with a centralized node.
9. The method of claim 1 , in which the network device is a serving cell.
10. The method of claim 1 , in which the network device is a neighbor cell.
11. The method of claim 1 , in which the network device is a centralized node.
12. The method of claim 1 , in which receiving the location comprises determining the location based on a serving cell reference signal received power (RSRP), a strongest beam from a serving cell, a serving cell precoder, a serving cell channel quality indicator (CQI), and/or a path loss estimate between the serving cell and the UE.
13. The method of claim 1 , in which receiving the location of the UE comprises receiving a geolocation of the UE.
14. The method of claim 1 , in which the information associated with the neighbor cell transmit beams comprises a precoder index with a known codebook and/or precoder weights.
15. The method of claim 1 , further comprising identifying undesired neighbor cell downlink transmit beams according to criteria of the interference-based distribution exceeding a predetermined threshold.
16. The method of claim 1 , in which the interference-based distribution comprises a signal to interference plus noise (SINR) distribution.
17. The method of claim 1 , in which the interference-based distribution comprises an interference over thermal (IoT) distribution.
18. An apparatus for wireless communication by a network device, comprising:
means for receiving a location of a user equipment (UE);
means for receiving information associated with neighbor cell transmit beams; and
means for predicting a spatial inter-cell interference-based distribution at the location of the UE based on the information associated with the neighbor cell transmit beams.
19. The apparatus of claim 18 , further comprising means for coordinating with the at least one neighbor cell to schedule downlink transmit beams in time and frequency resources different than time and frequency resources serving the UE.
20. The apparatus of claim 18 , further comprising means for training a neural network to learn the spatial inter-cell interference-based distribution for a plurality of locations.
21. The apparatus of claim 20 , in which the network device comprises a serving cell to learn the spatial inter-cell interference-based distribution is based on receiving channel state information (CSI) reference signal (CSI-RS) measurement reports from UEs at the plurality of locations.
22. A network device, comprising:
a processor;
a memory coupled with the processor;
instructions stored in the memory and operable, when executed by the processor, to cause the network device:
to receive a location of a user equipment (UE),
to receive information associated with neighbor cell transmit beams, and
to predict a spatial inter-cell interference-based distribution at the location of the UE based on the information associated with the neighbor cell transmit beams.
23. The network device of claim 22 , in which the instructions further cause the network device to train a neural network to learn the spatial inter-cell interference-based distribution for a plurality of locations.
24. The network device of claim 23 , in which the network device comprises a serving cell and learning the spatial inter-cell interference-based distribution is based on receiving channel state information (CSI) reference signal (CSI-RS) measurement reports from UEs at the plurality of locations.
25. The network device of claim 23 , in which the network device comprises a neighbor cell or a centralized node and learning the spatial inter-cell interference-based distribution is based on information received from a serving cell.
26. The network device of claim 23 , in which the network device comprises a serving cell and the instructions further the serving cell to train the neural network at the serving cell to compute neural network weights.
27. The network device of claim 22 , in which the network device comprises a serving cell, a neighbor cell, and/or a centralized node.
28. A non-transitory computer-readable medium having program code recorded thereon, the program code executed by a processor and comprising:
program code to receive a location of a user equipment (UE);
program code to receive information associated with neighbor cell transmit beams; and
program code to predict, by a network device, a spatial inter-cell interference-based distribution at the location of the UE based on the information associated with the neighbor cell transmit beams.
29. The non-transitory computer-readable medium of claim 28 , further comprising program code to train a neural network to learn the spatial inter-cell interference-based distribution for a plurality of locations.
30. The non-transitory computer-readable medium of claim 29 , in which the network device comprises a serving cell to learn the spatial inter-cell interference-based distribution is based on program code to receive channel state information (CSI) reference signal (CSI-RS) measurement reports from UEs at the plurality of locations.