Method performed by a radio network node for determining a changed bandwidth interval

Patent metadata

Field	Value
Publication number	US-12490195-B2
Application number	US-202018017521-A
Country	US
Kind code	B2
Filing date	Jul 27, 2020
Priority date	Jul 27, 2020
Publication date	Dec 2, 2025
Grant date	Dec 2, 2025

Abstract

Embodiments herein may relate to a method performed by a radio network node for handling communication over a bandwidth interval for one or more UEs, in a wireless communication network. The radio network node determines to change a currently used bandwidth interval based on a current state of usage of resources in the 5 currently used bandwidth interval related to number of UEs served and/or a current state of positions of the number of UEs served. With the proviso that it is determined to change the currently used bandwidth interval, the radio network node calculates a changed bandwidth interval to use. The radio network node then adjusts the currently used bandwidth interval to the calculated bandwidth interval; and uses the adjusted 10 bandwidth interval for communicating with the one or more UEs.

First claim

The invention claimed is:

1 . A method performed by a radio network node in a wireless communication network, the method comprising:
determining to change a currently used bandwidth interval used by the radio network node for communicating with one or more user equipments (UEs), referred to as served UEs, wherein the determination is based on at least one of: positions of the served UEs, or resource usage in the currently used bandwidth interval; adjusting the currently used bandwidth interval to an adjusted bandwidth interval; and using the adjusted bandwidth interval for communicating with the served UEs; wherein determining to change the currently used bandwidth interval is further based on at least one of: a prediction of resource usage corresponding to increasing or decreasing the currently used bandwidth interval, or a prediction of positions of the served UE.

2 . The method according to claim 1 , wherein using the adjusted bandwidth interval for communicating with the served UEs comprises reassigning the served UEs to respective resources in the adjusted bandwidth interval, based on at least one of: the number of served UEs and proximities of the served UEs to each other, as said positions of the served UEs.

3 . The method according to the claim 2 , wherein reassigning the served UEs to respective resources in the adjusted bandwidth interval comprises sending a command to a baseband unit to compact resources used by the served UEs within the adjusted bandwidth interval.

4 . The method according to claim 1 , wherein the determination is performed using a reinforcement learning process and is based on power savings that would be obtained from reducing the currently used bandwidth interval and a resulting quality of service (QoS) impact.

5 . The method according to claim 4 , further comprising monitoring the positions of the served UEs, the resource usage in the currently used bandwidth interval, a power consumption associated with use of the currently used bandwidth interval, QoS of the served UEs, and providing corresponding monitoring information to the reinforcement learning process.

6 . The method according to claim 4 , wherein the reinforcement learning process is implemented as a machine learning model, wherein for the current state(S), a changed bandwidth (A) is suggested, and wherein learning inputs are the current state S, the changed bandwidth A, and a parameter (R), where the parameter R is a profit having taken the changed bandwidth, A, in the current state, S.

7 . The method according to claim 6 , wherein the parameter R is modelled as the profit accrued based on a combination of power savings resulting from reducing the currently used bandwidth interval and the resulting QoS impact.

8 . The method according to claim 4 , wherein the reinforcement learning process is performed based on past decisions taken and updated during current operations.

9 . A method performed by a radio network node in a wireless communication network, the method comprising:
determining to change a currently used bandwidth interval used by the radio network node for communicating with one or more user equipments (UEs), referred to as served UEs, wherein the determination is based on at least one of: positions of the served UEs, or resource usage in the currently used bandwidth interval; adjusting the currently used bandwidth interval to an adjusted bandwidth interval; and using the adjusted bandwidth interval for communicating with the served UEs; wherein the determination is performed using an outer machine learning (ML) model using as inputs:
current number of served UEs;
information about positions of the served UEs, including proximities of served UEs to each other;
the currently used bandwidth interval;
predictions of one or both a change in the number of served UEs or changes in the positions of the served UEs; and
wherein the outer ML model outputs a decision to reduce or increase the currently used bandwidth interval, and by how much.

10 . The method according to claim 9 , wherein the adjusted bandwidth interval is one of the following 1.4, 3, 5, 10, 15, or 20 MHz.

11 . A radio network node for operation in a wireless communication network, wherein the radio network node comprises:
a communication interface configured for communicating with one or more user equipments (UEs), referred to as served UEs; and processing circuitry operatively associated with the communication interface and configured to:
determine to change a currently used bandwidth interval, which is used for communicating with the served UEs, wherein the determination is based on at least one of: positions of the served UEs, or resource usage in the currently used bandwidth interval;
adjust the currently used bandwidth interval to an adjusted bandwidth interval; and
use the adjusted bandwidth interval for communicating with the served UEs;
wherein the processing circuitry is configured to determine to change the currently used bandwidth interval further based on at least one of: a prediction of resource usage corresponding to increasing or decreasing the currently used bandwidth interval, or a prediction of positions of the served UEs.

12 . The radio network node according to claim 11 , wherein, with respect to using the adjusted bandwidth interval for communicating with the served UEs, the processing circuitry is configured to reassign the served UEs to respective resources in the adjusted bandwidth interval, based on at least one of: the number of served UEs, and proximities of the served UEs to each other, as said positions of the served UEs.

13 . The radio network node according to the claim 12 , wherein, to reassign the served UEs to respective resources in the adjusted bandwidth interval, the processing circuitry is configured to send a command to a baseband unit to compact resources used by the served UEs within the adjusted bandwidth interval.

14 . The radio network node according to claim 11 , wherein the determination is performed using a reinforcement learning process and is based on power savings that would be obtained by reducing the currently used bandwidth and a resulting quality of service (QoS) impact.

15 . The radio network node according to claim 14 , wherein the processing circuitry is configured to monitor the positions of the served UEs, the resource usage in the currently used bandwidth interval, a power consumption associated with use of the currently used bandwidth interval, and QoS of the served UEs, and provide corresponding monitoring information to the reinforcement learning process.

16 . The radio network node according to claim 14 , wherein the reinforcement learning process is implemented as an machine learning model, wherein for the current state(S), a changed bandwidth (A) is suggested, and wherein learning inputs are the current state S, the changed bandwidth A, and a parameter (R), where the parameter R is a profit having taken the changed bandwidth, A, in the current state, S.

17 . The radio network node according to claim 16 , wherein the parameter R is modelled as the profit accrued based on a combination of power savings resulting from reducing the currently used bandwidth and the resulting QoS impact.

18 . A radio network node for operation in a wireless communication network, wherein the radio network node comprises:
a communication interface configured for communicating with one or more user equipments (UEs), referred to as served UEs; and processing circuitry operatively associated with the communica

Assignees

• Ericsson Telefon Ab L M

Inventors

• Eleftheriadis Lackis
• Badrinath Ramamurthy
• Inam Rafia
• Nyström Cecilia
• Nikou Alexandros

Classifications

• H04W72/52

  based on load · CPC title

• H04L47/83

  based on usage prediction · CPC title

• G06N3/004

  Artificial life, i.e. computing arrangements simulating life · CPC title

• G06N20/00

  Machine learning · CPC title

• H04W24/02

  Arrangements for optimising operational condition · CPC title