Technique for predicting radio quality

Patent metadata

Field	Value
Publication number	US-2023361897-A1
Application number	US-202018023530-A
Country	US
Kind code	A1
Filing date	Aug 28, 2020
Priority date	Aug 28, 2020
Publication date	Nov 9, 2023
Grant date	—

Abstract

A technique for predicting radio quality in a wireless communication network depending on assumed positions of one or more base stations in an area to be covered by the wireless communication network is disclosed. A method implementation of the technique is performed by a computing unit and comprises the steps of determining, for a selected position in the area with respect to assumed positions of the one or more base stations, blocking object features indicative of a spatial pattern of blocking objects present in fields of view between the selected position and the assumed positions of the one or more base stations, and determining, based on the determined blocking object features, a predicted radio quality at the selected position using a machine learning model trained to map blocking object features for selected positions with respect to one or more base station positions to corresponding radio qualities at the selected positions.

First claim

1 . A method for predicting radio quality in a wireless communication network depending on assumed positions of one or more base stations in an area to be covered by the wireless communication network, the method being performed by a computing unit and comprising the steps of:
(a) determining, for a selected position in the area with respect to assumed positions of the one or more base stations, blocking object features indicative of a spatial pattern of blocking objects present in fields of view between the selected position and the assumed positions of the one or more base stations; and (b) determining, based on the determined blocking object features, a predicted radio quality at the selected position using a machine learning model trained to map blocking object features for selected positions with respect to one or more base station positions to corresponding radio qualities at the selected positions.

2 . The method of claim 1 , wherein steps (a) and (b) are performed for a plurality of different selected positions in the area with respect to the assumed positions of the one or more base stations to thereby obtain predicted radio qualities at the different selected positions, wherein the method further comprises:
(c) evaluating, based on the predicted radio qualities, a placement of the one or more base stations in the wireless communication network at the assumed positions.

3 . The method of claim 2 , wherein steps (a) to (c) are performed for a plurality of different assumed positions of the one or more base stations in the area to thereby obtain evaluations of different placements of the one or more base stations in the wireless communication network, wherein the method further comprises:
(d) selecting, based on the evaluations, a placement of the one or more base stations for deployment of the one or more base stations in the wireless communication network.

4 . The method of claim 1 , wherein the spatial pattern of the blocking objects is determined from a 3-dimensional environment scan performed in the area.

5 . The method of claim 4 , wherein determining the spatial pattern of the blocking objects comprises:
taking cross sections of the fields of view at different distances between the selected position and the assumed positions of the one or more base stations; and projecting blocking objects present in the fields of view at a particular distance onto the cross section at the particular distance.

6 . The method of claim 4 , wherein determining the spatial pattern of the blocking objects comprises calculating, for a cross section at a particular distance, a percentage of the cross section being blocked by blocking objects at the particular distance.

7 . The method of claim 1 , wherein the determined blocking object features are further indicative of at least one of absorption and reflection properties of the blocking objects present in the fields of view between the selected position and the assumed positions of the one or more base stations.

8 . The method of claim 7 , wherein at least one of the absorption and reflection properties of a blocking object is determined using an image recognition model trained with common blocking objects and at least one of their absorption and reflection properties.

9 . The method of claim 8 , wherein determining the at least one of the absorption and reflection properties of a blocking object is assisted by a manual marking of the blocking object input by a user on an image showing a field of view between the selected position and the assumed positions of the one or more base stations.

10 . The method of claim 7 , wherein at least one of the absorption and reflection properties of a blocking object is determined from radio measurements performed between a radio transmitter and a radio receiver when being placed iteratively at different positions around the blocking object.

11 . The method of claim 1 , further comprising:
identifying, among blocking objects detected in a 3-dimensional environment scan performed in the area, moveable blocking objects capable of moving in the area; and simulating movement of the movable blocking objects, wherein, in order to obtain a predicted radio quality at a selected position, steps (a) and (b) are repeatedly performed for different simulated movements of the movable blocking objects.

12 . The method of claim 1 , wherein the machine learning model is trained using training data comprising a plurality of sets of blocking object features as input data, each set of blocking object features associated with a different position of a wireless device with respect to one or more base stations of a representative wireless communication network, and corresponding radio qualities measured by the wireless device at the different positions as output data.

13 . The method of claim 12 , wherein, among the plurality of sets of blocking object features, each set of blocking object features is indicative of a spatial pattern of blocking objects present in fields of view between the current position of the wireless device and the one or more base stations.

14 . The method of claim 12 , wherein the training data is generated by moving the wireless device in an area of the representative wireless communication network and determining the sets of blocking object features and the corresponding radio qualities at different positions when moving the wireless device in the area.

15 . The method of claim 14 , wherein the input data of the training data further comprises, for each set of blocking features determined for a respective position when moving the wireless device in the area, distance information indicative of distances between the wireless device at the respective position and the one or more base stations.

16 . The method of claim 12 , wherein, for a set of blocking object features, the spatial pattern of the blocking objects is determined from a 3-dimensional environment scan performed in the area of the representative wireless communication network.

17 . The method of claim 16 , wherein determining the spatial pattern of the blocking objects comprises:
taking cross sections of the fields of view at different distances between the current position of the wireless device and the one or more base stations; and projecting blocking objects present in the fields of view at a particular distance onto the cross section at the particular distance.

18 . The method of claim 16 , wherein determining the spatial pattern of the blocking objects comprises calculating, for a cross section at a particular distance, a percentage of the cross section that is blocked by blocking objects at the particular distance.

19 - 23 . (canceled)

24 . A method for training a machine learning model configured to predict radio quality in a wireless communication network depending on assumed positions of one or more base stations in an area to be covered by the wireless communication network, the method being performed by a computing unit and comprising:
training the machine learning model using training data comprising a plurality of sets of blocking object features as input data, each set of blocking object features associated with a different position of a wireless device with respect to one or more base stations of a representative wireless communication network, and corresponding radio qualities measured by the wireless device at the different positions as output data, wherein, among the plurality of sets of blocking object features, each set of blocking o

Assignees

• Ericsson Telefon Ab L M

Inventors

• Rácz András
• Borsos Tamas
• Veres András
• Vaderna Peter

Classifications

• H04B17/373 Primary

  Predicting channel quality {or other radio frequency [RF]} parameters · CPC title

• H04W64/00

  Locating users or terminals {or network equipment} for network management purposes, e.g. mobility management · CPC title

• H04W4/02

  Services making use of location information · CPC title