US 12,363,539 B2
Method and system for detecting anomaly in radio access network
Jukka Pietarinen, Helsinki (FI); and Teemu Karhima, Helsinki (FI)
Assigned to Eilsa Oyj, Helsinki (FI)
Appl. No. 18/841,847
Filed by Elisa Oyj, Helsinki (FI)
PCT Filed Apr. 25, 2023, PCT No. PCT/FI2023/050225
§ 371(c)(1), (2) Date Aug. 27, 2024,
PCT Pub. No. WO2023/209280, PCT Pub. Date Nov. 2, 2023.
Claims priority of application No. 20225359 (FI), filed on Apr. 28, 2022.
Prior Publication US 2025/0071553 A1, Feb. 27, 2025
Int. Cl. H04W 12/122 (2021.01); H04W 24/08 (2009.01)
CPC H04W 12/122 (2021.01) [H04W 24/08 (2013.01)] 9 Claims
OG exemplary drawing
 
1. A method for detecting an anomaly in a radio access network (RAN), the method comprising:
receiving quality parameters data for a wide communication network established by a plurality of base stations associated with the RAN;
defining a first radio fingerprint based on the quality parameters data for the wide communication network, wherein defining the first radio fingerprint comprises:
isolating the quality parameters data for one or more base stations of the plurality of base stations installed in a geological area from the received quality parameters data for the wide communication network, wherein the quality parameters data is isolated by using a clustering algorithm; and
defining the first radio fingerprint based on the said isolated quality parameters data;
monitoring, by one of the one or more base stations of the plurality of base stations, installed in the geological area, quality parameters data for a local communication network, wherein the local communication network is a subset of the wide communication network;
defining a second radio fingerprint based on the quality parameters data for the local communication network;
comparing the first radio fingerprint and the second radio fingerprint to determine a variation between the quality parameters data for the wide communication network and the quality parameters data for the local communication network, wherein the variation between the quality parameters data for the wide communication network and the quality parameters data for the local communication network is determined by at least one of: binary segmentation algorithm, Pruned Exact Linear Time (PELT) algorithm, Z-score based algorithm; and
generating an alert signal indicative of a presence of a possible fake base station in the geological area based on the determined variation between the quality parameters data for the wide communication network and the quality parameters data for the local communication network.