US 12,464,390 B2
Clustering cell sites according to signaling behavior
Rensheng Zhang, Holmdel, NJ (US); Yaron Koral, Cherry Hill, NJ (US); and Arun Jotshi, Parsippany, NJ (US)
Assigned to AT&T Intellectual Property I, L.P., Atlanta, GA (US)
Filed by AT&T Intellectual Property I, L.P., Atlanta, GA (US)
Filed on Apr. 14, 2023, as Appl. No. 18/300,996.
Application 18/300,996 is a continuation of application No. 17/202,999, filed on Mar. 16, 2021, granted, now 11,653,234.
Prior Publication US 2023/0300647 A1, Sep. 21, 2023
This patent is subject to a terminal disclaimer.
Int. Cl. H04W 24/08 (2009.01); G06N 3/045 (2023.01); H04L 43/0823 (2022.01)
CPC H04W 24/08 (2013.01) [G06N 3/045 (2023.01); H04L 43/0823 (2013.01)] 20 Claims
OG exemplary drawing
 
1. A method, comprising:
classifying, by a system comprising a processor and using a neural network, respective cell devices, comprising a cell device, of a cluster of cell devices of respective clusters of cell devices, based on respective encoded reduced dimensionality vectors associated with the respective cell devices and information relating to the respective clusters of cell devices, wherein the respective encoded reduced dimensionality vectors are generated based on performing feature reforming of signal measurement data to generate respective frequency feature reduced dimensionality vectors and encoding the respective frequency feature reduced dimensionality vectors, wherein the signal measurement data is representative of signal measurements of signals associated with the respective clusters of cell devices, and wherein the neural network is determined to be usable to determine a representation of a cellular network, comprising the respective clusters of cell devices; and
determining, by the system using the neural network, whether an abnormal condition associated with the cell device has occurred based on the classifying of the cell device and based on a network security criterion that defines the abnormal condition.