US 11,943,640 B2
Technique for generating synthetic data for radio access network configuration recommendation
Yu Wang, Solna (SE); and Wenfeng Hu, Täby (SE)
Assigned to TELEFONAKTIEBOLAGET LM ERICSSON (PUBL), Stockholm (SE)
Appl. No. 17/610,280
Filed by Telefonaktiebolaget LM Ericsson (publ), Stockholm (SE)
PCT Filed May 28, 2019, PCT No. PCT/EP2019/063799
§ 371(c)(1), (2) Date Nov. 10, 2021,
PCT Pub. No. WO2020/239203, PCT Pub. Date Dec. 3, 2020.
Prior Publication US 2022/0240106 A1, Jul. 28, 2022
Int. Cl. H04W 24/02 (2009.01); G06N 3/045 (2023.01); G06N 3/088 (2023.01); H04W 16/22 (2009.01)
CPC H04W 24/02 (2013.01) [G06N 3/045 (2023.01); G06N 3/088 (2013.01); H04W 16/22 (2013.01)] 20 Claims
OG exemplary drawing
 
1. An apparatus for generating synthetic data as input for a machine learning process that recommends radio access network (RAN) configurations, the apparatus comprising:
processing circuitry;
memory containing instructions executable by the processing circuitry whereby the apparatus is operative to:
obtain a noise input;
generate, using a trained generative machine learning model, synthetic data from the noise input, wherein the generative machine learning model has been trained together with a discriminative machine learning model as adversaries based on non-synthetic data associating non-synthetic configuration management (CM) parameter values, non-synthetic RAN characteristic parameter values, and non-synthetic performance indicator values; wherein each non-synthetic performance indicator value indicates a performance for a given RAN configuration, as defined by one or more of the non-synthetic CM parameter values and a given RAN characteristic as defined by one or more of the non-synthetic RAN characteristic parameter values, wherein the synthetic data, in the same form as the non-synthetic data, comprises at least one of one or more synthetic CM parameter values, one or more synthetic RAN characteristic parameter values, and one or more synthetic performance indicator values; and
output the synthetic data for the machine learning process.