Netscout Systems Texas, LLC

Sunil Bhat

A method is provided of limiting machine learning (ML) model training data, including receiving a maximum number of entries per grid box N. Grids are defined, each grid having grid boxes and covering a geographic area defined by a different cell list of a communication network, each grid box covering a different portion of the geographic area. Truth call records having truth data reporting geolocation (GL) data are selected from multiple received call records. A grid box is determined for each truth call record that covers a geographic area including the GL indicated by the truth call record's GL data. Truth call records are selectively assigned to their determined grid box in a fashion to not exceed the maximum number of entries per grid box N, and the truth data and signal detail data of only the assigned truth call records are output as the ML model training data.

Reduction and geo-spatial distribution of training data for geolocation prediction using machine learning

A method includes receiving call records from a control plane, each call record including a cell list identifying the server cell for the UE call session at the time the call record was generated and an ordered set of neighbor cells, ordered based on a characteristic of signals from the neighbor cells. Call records having truth data are selected, wherein the truth data includes geolocation (GL) data reported to be a GL associated with the call record. GL data of the selected call records is stored in association with the cell list for the selected call records. A centroid is determined as a function of the GL data associated with each of the selected call records that includes the associated cell list. The centroids are stored in association with the corresponding cell list, and can be retrieved as a prediction for a GL based on submission of a cell list.

Growing community of inventors

Plano, TX, United States of America

Sunil Bhat