Transfer learning of network traffic prediction model among cellular base stations

Patent metadata

Field	Value
Publication number	US-11696153-B2
Application number	US-202117391708-A
Country	US
Kind code	B2
Filing date	Aug 2, 2021
Priority date	Aug 13, 2020
Publication date	Jul 4, 2023
Grant date	Jul 4, 2023

Abstract

Transfer learning based on prediction determines a similarity between a source base station and a target base station. Importance of parameters is determined and training is adjusted to respect the importance of parameters. A lack of historical data is compensated by selecting a base station as source base station which has a larger amount of historical data.

First claim

What is claimed is:

1. A server configured to manage traffic prediction model transfer learning among 5G base stations, the server comprising:
one or more processors; and
one or more memories, the one or more memories storing a program, wherein execution of the program by the one or more processors is configured to cause the server to at least:
receive a first plurality of base station statistics, wherein the first plurality of base station statistics includes a first data set of a first size from a first base station;
receive a second plurality of base station statistics, wherein the second plurality of base station statistics includes a second data set of a second size corresponding to a second base station;
select the first base station as a source base station;
train a similarity network;
receive a source prediction model from the first base station and an importance score matrix;
receive a prediction model request from a target base station, wherein the target base station is the second base station;
compute a first similarity using the similarity network;
obtain a first scaled importance score matrix based on the importance score matrix and based on the first similarity; and
send the source prediction model and the first scaled importance score matrix to the second base station,
whereby the second base station is configured to use the source prediction model, and the first scaled importance score matrix to generate a target prediction model and predict radio system parameters relevant to the second base station, wherein the radio system parameters include a future value of user data traffic passing through the second base station.

2. The server of claim 1 , wherein execution of the program by the one or more processors is further configured to cause the server to:
receive a third data set of a third size from a third base station;
determine a second similarity using the similarity network and the third data set;
compute a second scaled importance score matrix based on the second similarity and the importance score matrix; and
send the source prediction model and the second scaled importance score matrix to the third base station.

3. The server of claim 1 , wherein the first data set includes a histogram of a traffic history of the source base station, wherein an abscissa of the histogram is proportional to bits per second, the first data set further includes a first indication of frequency bands supported by the source base station, a second indication of radio access types supported by the source base station, a third indication of 5G class types supported by the source base station and a fourth indication of user density currently supported by the source base station, and wherein a first node vector is formed based on the first data set.

4. The server of claim 3 , wherein the importance score matrix is a second order derivative of a Fisher information matrix with respect to weights of the source prediction model, and the first scaled importance score matrix is a product of the importance score matrix and the first similarity.

5. The server of claim 1 , wherein execution of the program by the one or more processors is further configured to cause the server to select a candidate base station with a largest data set as the source base station.

6. The server of claim 3 , wherein the similarity network comprises an autoencoder, and execution of the program by the one or more processors is further configured to cause the server to train the similarity network by using gradient descent to update parameters of the autoencoder based on an autoencoder loss, wherein the autoencoder loss is a distance between the first node vector and an estimated node vector, wherein the estimated node vector is an output of the similarity network.

7. The server of claim 6 , wherein execution of the program by the one or more processors is further configured to cause the server to compute the first similarity by:
obtaining a second node vector from the target base station;
obtaining a first latent vector as a first output of the autoencoder when the first node vector is input to the autoencoder;
obtaining a second latent vector as a second output of the autoencoder when the second node vector is input to the autoencoder; and
computing the first similarity as a cosine similarity between the first latent vector and the second latent vector.

8. A method for managing traffic prediction model transfer learning among 5G base stations, the method comprising:
receiving a first plurality of base station statistics, wherein the first plurality of base station statistics includes a first data set of a first size from a first base station;
receiving a second plurality of base station statistics, wherein the second plurality of base station statistics includes a second data set of a second size corresponding to a second base station;
selecting the first base station as a source base station;
training a similarity network;
receiving a source prediction model from the first base station and an importance score matrix;
receiving a prediction model request from a target base station, wherein the target base station is the second base station;
computing a first similarity using the similarity network;
obtaining a first scaled importance score matrix based on the importance score matrix and based on the first similarity; and
sending the source prediction model and the first scaled importance score matrix to the second base station.

9. The method of claim 8 , further comprising:
receiving a third data set of a third size from a third base station;
determining a second similarity using the similarity network and the third data set;
computing a second scaled importance score matrix based on the second similarity and the importance score matrix; and
sending the source prediction model and the second scaled importance score matrix to the third base station.

10. The method of claim 8 , wherein the first data set includes a histogram of a traffic history of the source base station, wherein an abscissa of the histogram is proportional to bits per second, the first data set further includes a first indication of frequency bands supported by the source base station, a second indication of radio access types supported by the source base station, a third indication of 5G class types supported by the source base station and a fourth indication of user density currently supported by the source base station, and wherein a first node vector is formed based on the first data set.

11. The method of claim 10 , wherein the similarity network comprises an autoencoder, and the training the similarity network comprises using gradient descent to update parameters of the autoencoder based on an autoencoder loss, wherein the autoencoder loss is a distance between the first node vector and an estimated node vector, wherein the estimated node vector is an output of the similarity network.

12. The method of claim 11 , wherein the computing the first similarity comprises:
obtaining a second node vector from the target base station;
obtaining a first latent vector as a first output of the autoencoder when the first node vector is input to the autoencoder;
obtaining a second latent vector as a second output of the autoencoder when the second node vector is input to the autoencoder; and
computing the first similarity as a cosine similarity between the first latent vector and the second latent vector.

13. The method of claim 8 , further comprising selecting a candidate base station with a largest data set as the source base station.

14. The method of claim 8 ,

Assignees

• Samsung Electronics Co Ltd

Inventors

• Chen Xi
• Wang Ju
• Li Hang
• Xu yi tian
• Wu Di
• Liu Xue
• Dudek Gregory Lewis
• Lee Taeseop
• Park Intaik

Classifications

• H04W16/22 Primary

  Traffic simulation tools or models · CPC title

• H04W16/18

  Network planning tools · CPC title

• H04W16/04 Primary

  Traffic adaptive resource partitioning · CPC title

• H04W24/08

  Testing, {supervising or monitoring} using real traffic · CPC title

• G06N3/0455

  Auto-encoder networks; Encoder-decoder networks · CPC title