Telefonaktiebolaget Lm Ericsson (publ)

Wenfeng Hu

Yu Wang

Jaeseong Jeong

Selim Ickin

Farnaz Moradi

Martha Vlachou-Konchylaki

Tobias Ley

Saurabh Singh

Vladimir Verbulskii

Rodrigo Correia

Wei Huang

Michel Gokan Khan

Carolyn Cartwright

Huiyuan Wang

A method performed by a central server node in a distributed machine learning environment is provided. The method includes: managing distributed machine learning for a plurality of local client nodes, such that a first set of the plurality of local client nodes are assigned to assist training of a first central model and a second set of the plurality of local client nodes are assigned to assist training of a second central model; obtaining information regarding network conditions for the plurality of local client nodes; clustering the plurality of local client nodes into one or more clusters based at least in part on the information regarding network conditions; re-assigning a local client node in the first set to the second set based on the clustering; and sending to the local client node a message including model weights for the second central model.

Distributed machine learning using network measurements

A network metrics repository stores cell performance metrics and rule-based data measured during operation of a communication network. A policy neural network circuit has an input layer having input nodes, a sequence of hidden layers, and at least one output node. A processor trains the policy neural network circuit to approximate a baseline rule-based policy for controlling a tilt angle of a remote electrical tilt (RET) antenna based on the rule-based data. The processor provides a live cell performance metric to input nodes, adapts weights that are used by the input nodes responsive to output of the output node, and controls operation of the tilt angle of the RET antenna based on the output The output node provides the output responsive to processing a stream of cell performance metrics through the input nodes. The processor controls operation of the RET antenna based on the output.

Neural network circuit remote electrical tilt antenna infrastructure management based on probability of actions

A method () for performance modeling of a plurality of microservices () includes deploying the plurality of microservices () within a network (). The plurality of microservices () are communicatively coupled to generate at least one service chain () for providing at least one service. Based on a resource allocation configuration, an initial set of training data for the plurality of microservices within the network () is determined. At least a portion of data is excluded from the initial set of training data to generate a subset of training data. A Quality of Service (QoS) behaviour model is generated based on the subset of the training data.

Performance modeling for cloud applications

A technique for generating synthetic data as input for a machine learning process that recommends radio access network, RAN, configurations is presented. An apparatus implementation is configured to generate synthetic data from a noise input, using a trained generative machine learning model, wherein the generative machine learning model has been trained together with a discriminative machine learning model as adversaries based on non-synthetic data. The non-synthetic data comprises non-synthetic configuration management, CM, parameter values, non-synthetic RAN characteristic parameter values and non-synthetic performance indicator values. The synthetic data is in the same form as the non-synthetic data and comprises synthetic configuration management, CM, parameter values, synthetic RAN characteristic parameter values and synthetic performance indicator values. The apparatus is also configured to output the synthetic data for the machine learning process.

Technique for generating synthetic data for radio access network configuration recommendation

A pre-training apparatus and method for reinforcement learning based on a Generative Adversarial Network (GAN) is provided. GAN includes a generator and a discriminator. The method comprising receiving training data from a real environment where the training data includes a data slice corresponding to a first state-reward pair and a first state-action pair, training the GAN using the training data, training a relations network to extract a latent relationship of the first state-action pair with the first state-reward pair in a reinforcement learning context, causing the generator trained with training data to generate first synthetic data, processing a portion of the first synthetic data in the relations network to generate a resulting data slice, merging the second state-action pair portion of the first synthetic data with the second state-reward pair from the relations network to generate second synthetic data to update a policy for interaction with the real environment.

Pre-training system for self-learning agent in virtualized environment

Growing community of inventors

Täby, Sweden

Wenfeng Hu