US 12,132,627 B1
Machine learning model for assessing network performance
Yinjun Zhu, Sammamish, WA (US); Jan Olav Opedal, Ellensburg, WA (US); Suchit Satpathy, Hackettstown, NJ (US); Srikrishna Srinivasan, Sammamish, WA (US); and Yanbing Su, Frisco, TX (US)
Assigned to T-Mobile USA, Inc., Bellevue, WA (US)
Filed by T-Mobile USA, Inc., Bellevue, WA (US)
Filed on Aug. 4, 2023, as Appl. No. 18/365,814.
Int. Cl. G06F 15/173 (2006.01); H04L 41/16 (2022.01); H04L 43/06 (2022.01); H04L 43/08 (2022.01); H04L 65/1016 (2022.01)
CPC H04L 43/08 (2013.01) [H04L 41/16 (2013.01); H04L 43/06 (2013.01); H04L 65/1016 (2013.01)] 20 Claims
OG exemplary drawing
 
1. A system for assessing network performance during Internet protocol multimedia subsystem (IMS) data sessions, the system comprising:
one or more processors; and
one or more non-transitory, computer-readable media having instructions recorded thereon that when executed by the one or more processors cause operations comprising:
receiving real-time telecommunications network data,
wherein the real-time telecommunications network data includes data measured by a network element at each of a plurality of measurement intervals during an IMS data session, and
wherein the IMS data session is conducted over a telecommunications network that includes the network element;
inputting to a trained model, the real-time telecommunications network data;
computing, using the trained model, a subjective quality rating for each of the plurality of measurement intervals by:
identifying attributes in the real-time telecommunications network data, wherein the attributes include raw data at each of the plurality of measurement intervals during the IMS data session; and
using the attributes as well as data from one or more similar IMS data sessions to determine a subjective quality rating;
receiving, as output from the trained model, the subjective quality rating for each of the plurality of measurement intervals;
generating a performance score by aggregating the output from the trained model from each of the plurality of measurement intervals during the IMS data session;
generating a report including the performance score; and
modifying an attribute of the telecommunications network based on the report.