US 11,658,880 B2
Transfer learning for radio resource management
Euhanna Ghadimi, Bromma (SE); Jonas Wiorek, Stockholm (SE); and Sholeh Yasini, Sundbyberg (SE)
Assigned to TELEFONAKTIEBOLAGET LM ERICSSON (PUBL), Stockholm (SE)
Appl. No. 17/434,352
Filed by Telefonaktiebolaget LM Ericsson (publ), Stockholm (SE)
PCT Filed Feb. 27, 2019, PCT No. PCT/IB2019/051584
§ 371(c)(1), (2) Date Aug. 26, 2021,
PCT Pub. No. WO2020/174262, PCT Pub. Date Sep. 3, 2020.
Prior Publication US 2022/0158907 A1, May 19, 2022
Int. Cl. H04L 29/08 (2006.01); H04L 41/16 (2022.01); H04L 41/044 (2022.01)
CPC H04L 41/16 (2013.01) [H04L 41/044 (2013.01)] 20 Claims
OG exemplary drawing
 
1. A network operator device performing as a source for policies related to radio resource management, RRM, in a radio access network, RAN, the network operator device comprising:
a communication interface configured to intermediate data exchange with a target learner in the RAN; and
a data processing unit connected to the communication interface and configured to prepare and supply a source policy, SP, for an RRM-related task, via the communication interface to the target learner, wherein the target learner is enabled to provide a production policy, PP, to an actor that determines an RRM-related action for accomplishing the RRM-related task, and wherein:
the SP is prepared based on knowledge related to a first cell having a first cell type, and the target learner is enabled to provide the PP related to a second cell having a second cell type;
the SP is prepared using data from multiple cells, and the target learner is enabled to provide the PP from a specific cell;
the SP is prepared using simulated data, and the target learner is enabled to provide the PP which is configured to use real network data;
the SP is prepared using a high volume of correlated low-cost sources of data, and the target learner is enabled to provide the PP configured to use a lower volume of data, which is more precise and expensive than the high volume of correlated low-cost sources of data;
the SP is prepared using offline available data, and the target learner is enabled to provide the PP configured to use online network data; and/or
the SP is prepared using 4G data or for a 4G feature, and the target learner is enabled to provided the PP configured for a 5G feature.