Method and apparatus for monitoring a telecommunication network

Patent metadata

Field	Value
Publication number	US-11146344-B2
Application number	US-201916259069-A
Country	US
Kind code	B2
Filing date	Jan 28, 2019
Priority date	Feb 16, 2018
Publication date	Oct 12, 2021
Grant date	Oct 12, 2021

Abstract

Embodiments relate to an apparatus for monitoring a telecommunication network including one or more telecommunication channels. The apparatus includes a processor configured after executing computer code to design a convolutional neural network configured for determining an impairment type of a telecommunication channel as a function of a channel frequency response of the telecommunication channel, by selecting at least one of a number of convolutional layers, a number of filters for respective convolutional layers, and/or a size of filters for respective convolutional layers; and train the convolutional neural network based on training data specifying, for respective telecommunication channels, a channel frequency response and an associated impairment type.

First claim

The invention claimed is:

1. An apparatus for monitoring a telecommunication network including one or more telecommunication channels, the apparatus comprising:
a memory storing computer code; and
at least one processor configured to execute the computer code which configures the at least one processor to:
design a convolutional neural network configured to determine an impairment type of at least one telecommunication channel of the one or more telecommunication channels as a function of a channel frequency response of the at least one telecommunication channel, the designing the convolutional neural network including selecting at least one of a number of convolutional layers, a number of filters for respective convolutional layers, and/or a size of filters for the respective convolutional layers, and
train the convolutional neural network based on training data specifying, for one or more respective telecommunication channels, a channel frequency response and an associated impairment type.

2. The apparatus according to claim 1 , wherein
the one or more telecommunication channels are a plurality of telecommunication channels; and
the designing the convolutional neural network comprises:
selecting at least one of the number of convolutional layers, the number of filters and/or the size of filters as a function of analysis results of channel frequency responses for the plurality of telecommunication channels.

3. The apparatus according to claim 2 , wherein the computer code further configures the at least one processor to:
determine a fundamental frequency for respective channel frequency responses,
determine at least one fundamental frequency distribution based on the determined fundamental frequencies, and
select the number of convolutional layers as a function of the at least one fundamental frequency distribution.

4. The apparatus according to claim 3 , wherein the computer code further configures the at least one processor to:
determine a lag for respective channel frequency responses,
determine at least one lag distribution based on the determined fundamental frequencies, and
select the filter size as a function of the at least one lag distribution.

5. The apparatus according to claim 4 , wherein the computer code further configures the at least one processor to:
determine a YIN function of a channel frequency response; and
determine the fundamental frequency and/or the lag based on the YIN function.

6. The apparatus according to claim 1 , wherein
the one or more telecommunication channels comprises at least one digital subscriber line (DSL) line;
the channel frequency response is a Hlog; and
the convolutional neural network comprises a first convolutional layer comprising between 8 and 16 filters of size comprised between 4 and 12.

7. The apparatus according to claim 6 , wherein the convolutional neural network comprises at least a second convolutional layer.

8. The apparatus according to claim 1 , wherein the computer code further configures the at least one processor to
determine, with the trained convolutional neural network, an impairment type associated with a telecommunication channel, as a function of a channel frequency response of the telecommunication channel.

9. The apparatus according to claim 8 , wherein the computer code further configures the at least one processor to:
control a change of configuration of an access node and/or of a client device associated with the telecommunication channel, based on the determined impairment type.

10. A method for monitoring a telecommunication network including one or more telecommunication channels, the method comprising:
designing a convolutional neural network configured for determining an impairment type of at least one telecommunication channel of the one or more telecommunication channels as a function of a channel frequency response of the at least one telecommunication channel, the designing the convolutional neural network including selecting at least one of a number of convolutional layers, a number of filters for respective convolutional layers, and/or a size of filters for the respective convolutional layers, and
training the convolutional neural network based on training data specifying, for one or more respective telecommunication channels, a channel frequency response and an associated impairment type.

11. A non-transitory computer-readable medium storing computer-executable instructions which when executed by a computer causes the computer to perform the method of claim 10 .

Assignees

• Nokia Solutions & Networks Oy

Inventors

• Dupuis Nicolas
• Van Damme Axel

Classifications

• G06N3/045

  Combinations of networks · CPC title

• G06N3/08

  Learning methods · CPC title

• H04L41/16 Primary

  using machine learning or artificial intelligence · CPC title

• G06N3/09

  Supervised learning · CPC title

• G06N3/0464

  Convolutional networks [CNN, ConvNet] · CPC title