T-Mobile Innovations LLC

Hui-Lin Chang

Lance Paul Lukens

Daqi Li

Shawn Derek Wallace

Piradee Nganrungruang

Wayne Alexander Thompson

Leland John Garris

Farhana Sharmin Munnee

Diksha Agarwal

Systems and methods for making telecommunication network predictions on a granular geographic area-scale. The method can include receiving network key performance indicators (KPIs) associated with a plurality of user equipment (UEs) and then parsing the network KPIs into a plurality of geographic area-based categories corresponding to network KPIs obtained by cell towers or UEs located within a given geographic area, like a zip code. The method can further include combining the geographic area-based categories into clusters determined based on one or more similarities in the network KPIs of each, and predicting a most-influential one of the network KPIs within at least one of the clusters via a machine learning (ML) model. In response, the method can then cause an indicator to be generated at a user interface, the indicator indicating the most-influential one of the network KPIs within the at least one of the clusters.

Telecommunications network predictions based on machine learning using aggregated network key performance indicators

Anomalies in network traffic are detected using machine learning. A plurality of machine learning models is employed to determine whether there are anomalies in network traffic of an MPLS (Multiprotocol Label Switching) network that can affect the performance of devices in the network. A first machine learning model is trained on network traffic passed through network tunnels of a plurality of routers in the network. A second machine learning model is trained on router-specific network traffic passed through router-specific network traffic for a subset of the network tunnels associated with a particular router. The first machine learning model is employed to determine a network anomaly, and the second machine learning model is employed to determine a router-specific anomaly. A router error is identified when both a network anomaly and a router-specific anomaly are determined. An indication of the router error is communicated to a computing device.

Network anomaly detection using machine learning models

A cell site maintenance system. The system comprises a plurality of data stores, a processor, a memory coupled to the processor, and a part replacement prediction application stored in memory. When executed by the processor the application extracts features based on data read from the data stores, executes a plurality of part machine learning models, where each part machine learning model analyses some of the extracted features by a multi-label classification model trained for a specific part associated with that part machine learning model and each part machine learning model outputs a probability that the associated part can fix the cell site associated with the trouble ticket, and, based on a probability output by a part machine learning model associated with a first part exceeding a predefined threshold, determining that the first part is to be pulled from a part inventory to make ready for a service truck.

Cell site repair part prediction machine learning system

Growing community of inventors

Brambleton, VA, United States of America

Hui-Lin Chang