LM Ericsson Machine Learning Patent Granted

ChangeBridge: Patent Grants - AI & Computing (G06N)

Published March 24th, 2026

Detected March 25th, 2026

Summary

The USPTO has granted a patent (US12585986B2) to LM Ericsson for a machine learning method in communication networks. The patent covers training a model, serializing local model updates, and transmitting them to an aggregator entity.

View original document View source feed page

What changed

The United States Patent and Trademark Office (USPTO) has granted patent US12585986B2 to Telefonaktiebolaget LM Ericsson for a method related to machine learning in communication networks. The patent details a process where a network entity trains a model, creates a serialized representation of local model updates by removing structural information, and transmits this representation to an aggregator. This grant is a new intellectual property right, not a regulatory change impacting compliance obligations for other entities.

As this is a granted patent, there are no immediate compliance actions required for regulated entities. However, companies operating in the telecommunications and AI sectors, particularly those involved in federated learning or distributed model training, may wish to review the patent's claims to understand potential intellectual property landscapes and avoid infringement. The patent's effective date is March 24, 2026.

Source document (simplified)

← USPTO Patent Grants

Methods, apparatus and machine-readable media relating to machine-learning in a communication network

Grant US12585986B2 Kind: B2 Mar 24, 2026

Assignee

Telefonaktiebolaget LM Ericsson (publ)

Inventors

Karl Norrman, Martin Isaksson

Abstract

A method performed by a first network entity in a communications network includes training a model to obtain a local model update including an update to values of one or more parameters of the model, in which training the model includes inputting training data into a machine learning algorithm. The method further includes applying a serialisation function to the local model update to construct a serial representation of the local model update, thereby removing information indicative of a structure of the model, and transmitting the serial representation of the local model update to an aggregator entity in the communications network.