User scheduling using a graph neural network

Patent metadata

Field	Value
Publication number	US-12490288-B2
Application number	US-202117551257-A
Country	US
Kind code	B2
Filing date	Dec 15, 2021
Priority date	Dec 15, 2021
Publication date	Dec 2, 2025
Grant date	Dec 2, 2025

Abstract

The present disclosure relates to a device for use in a wireless network, the device including: a processor configured to: provide input data to a trained graph neural network model, the input data being indicative of a graph representation of a plurality of wireless communication devices, wherein the trained graph neural network model is configured to provide output data being indicative of a scheduled user set including the plurality of wireless communication devices; and instruct user scheduling of the plurality of wireless communication devices based on the output data of the trained graph neural network model.

First claim

What is claimed is:

1 . A device for use in a wireless network, the device comprising:
a processor configured to: provide input data to a trained graph neural network model, the input data being representative of a graph representation of a plurality of wireless communication devices, wherein the graph representation comprises a plurality of nodes associated with the plurality of wireless communication devices, and a plurality of edges associated with interactions among the plurality of wireless communication devices, wherein each node of the plurality of nodes has an input feature vector associated therewith, the input feature vector being representative of one or more features of the wireless communication device associated with the node, wherein the trained graph neural network model is configured to provide output data representative of a scheduled user set comprising the plurality of wireless communication devices; and instruct user scheduling of the plurality of wireless communication devices based on the output data of the trained graph neural network model.

2 . The device according to claim 1 ,
wherein the output data comprises a set of scores, and wherein each score of the set of scores is associated with a wireless communication device of the plurality of wireless communication devices.

3 . The device according to claim 1 ,
wherein the input data comprises correlation data representative of a correlation among the wireless communication devices of the plurality of wireless communication devices.

4 . The device according to claim 3 ,
wherein the correlation data comprises a pairwise signal-to-interference-plus-noise ratio among the wireless communication devices.

5 . The device according to claim 1 ,
wherein the plurality of nodes comprises, for each wireless communication device of the plurality of wireless communication devices, a number of nodes corresponding to a number of layers associated with the wireless communication device.

6 . The device according to claim 1 ,
wherein the graph representation comprises a respective graph representation for each sub-band of a resource block group of one or more resource block groups of the wireless network.

7 . The device according to claim 6 ,
wherein each graph representation comprises, for each wireless communication device of the plurality of wireless communication devices, a number of nodes corresponding to a number of layers associated with the wireless communication device in the respective sub-band.

8 . The device according to claim 1 ,
wherein the one or more features of the wireless communication device comprise one or more of: layer signal-to-interference-and-noise ratio, a user weight, a precoding vector, and/or combinations thereof.

9 . The device according to claim 1 ,
wherein each node of the plurality of nodes is connected with each other node of the plurality of nodes.

10 . The device according to claim 1 ,
wherein the graph representation further comprises a virtual node not associated with any wireless communication device of the plurality of wireless communication devices.

11 . The device according to claim 1 ,
wherein an edge between a node of the plurality of nodes and an other node of the plurality of nodes is associated with channel state information of the node and of the other node.

12 . The device according to claim 11 ,
wherein the channel state information comprises one or more of: single-user Multiple-Input Multiple-Output precoding matrix indicator per sub-band; single-user Multiple-Input Multiple-Output beamformer; channel quality indicator per layer per sub-band; and/or rank indicator.

13 . The device according to claim 1 ,
wherein the graph representation further comprises a plurality of attention scores, each attention score being representative of a weighted relationship between a node of the plurality of nodes and another node of the plurality of nodes.

14 . The device according to claim 1 ,
wherein the trained graph neural network model is configured as a xAPP in a radio access network intelligent controller, or wherein the processor is configured to access the trained graph neural network model in a distributed unit of a radio access network of the wireless network.

15 . A method of operating a wireless network, the method comprising:
providing input data to a trained graph neural network model, the input data being indicative of a graph representation of a plurality of wireless communication devices, wherein the graph representation comprises a plurality of nodes associated with the plurality of wireless communication devices, and a plurality of edges associated with interactions among the plurality of wireless communication devices, wherein each node of the plurality of nodes has an input feature vector associated therewith, the input feature vector being representative of one or more features of the wireless communication device associated with the node, wherein the trained graph neural network model is configured to provide output data representative of a scheduled user set comprising the plurality of wireless communication devices; and performing user scheduling of the plurality of wireless communication devices based on the output data of the trained graph neural network model.

16 . The method according to claim 15 ,
wherein the output data comprises a set of scores, wherein each score of the set of scores is associated with a wireless communication device of the plurality of wireless communication devices, and wherein performing user scheduling comprises allocating resources to the plurality of wireless communication devices in descending order of associated scores.

17 . A device for use in a wireless network, the device comprising:
a processor configured to: provide input data to a trained graph neural network model, the input data being representative of a graph representation of a plurality of wireless communication devices, wherein the graph representation comprises a plurality of nodes associated with the plurality of wireless communication devices, and a plurality of edges associated with interactions among the plurality of wireless communication devices, wherein the input data comprises correlation data representative of a correlation among the wireless communication devices of the plurality of wireless communication devices, wherein the correlation data comprises a pairwise signal-to-interference-plus-noise ratio among the wireless communication devices, wherein the trained graph neural network model is configured to provide output data representative of a scheduled user set comprising the plurality of wireless communication devices; and instruct user scheduling of the plurality of wireless communication devices based on the output data of the trained graph neural network model.

Assignees

• Intel Corp

Inventors

• Narasimha Swamy Vasuki
• Orhan Oner
• Nikopour Hosein
• Talwar Shilpa

Classifications

• G06N3/08

  Learning methods · CPC title

• H04B7/0452

  Multi-user MIMO systems · CPC title

• G06N3/084

  Backpropagation, e.g. using gradient descent · CPC title

• G06N7/01

  Probabilistic graphical models, e.g. probabilistic networks · CPC title

• G06N3/044

  Recurrent networks, e.g. Hopfield networks · CPC title