中科微点-全球专利检索

  1. Login
  2. Sign Up

Search Results

19 records (took 0.02 seconds)

    TRAFFIC MATRIX PREDICTION AND FAST REROUTE PATH COMPUTATION IN PACKET NETWORKS
    11.
    发明申请
    TRAFFIC MATRIX PREDICTION AND FAST REROUTE PATH COMPUTATION IN PACKET NETWORKS 有权

    公开(公告)号：US20210029030A1

    公开(公告)日：2021-01-28

    申请号：US17068786

    申请日：2020-10-12

    Applicant: AT&T Intellectual Property I, L.P.

    Inventor： Gagan Choudhury ,  Gaurav Thakur ,  Simon Tse ,  Kathleen Meier-Hellstern

    IPC: H04L12/721 ,  H04L12/715 ,  H04L12/26 ,  H04Q11/00 ,  H04L12/723

    Abstract: A processing system including at least one processor may obtain traffic measurements for end-to-end paths in a telecommunication network, calculate traffic estimates for the end-to-end paths in future time periods based on the traffic measurements in accordance with at least one machine learning model, calculate traffic estimates for primary paths in the telecommunication network based upon the traffic estimates for the end-to-end paths, compute a backup path configuration for a primary path of the telecommunication network for the future time periods based upon the traffic estimates for the primary paths in the future time periods, detect a change in the backup path configuration for the primary path in a future time period based upon the computing, and adjust a backup path in accordance with the backup path configuration when the change in the backup path configuration is detected.

    MACHINE LEARNING TECHNIQUES FOR SELECTING PATHS IN MULTI-VENDOR RECONFIGURABLE OPTICAL ADD/DROP MULTIPLEXER NETWORKS
    12.
    发明申请
    MACHINE LEARNING TECHNIQUES FOR SELECTING PATHS IN MULTI-VENDOR RECONFIGURABLE OPTICAL ADD/DROP MULTIPLEXER NETWORKS 审中-公开

    公开(公告)号：US20200092026A1

    公开(公告)日：2020-03-19

    申请号：US16135844

    申请日：2018-09-19

    Applicant: AT&T Intellectual Property I, L.P.

    Inventor： Martin Birk ,  David Frederick Lynch ,  Gaurav Thakur ,  Simon Tse

    IPC: H04J14/02 ,  H04B10/079 ,  H04L12/751

    Abstract: Devices, computer-readable media and methods are disclosed for selecting paths in reconfigurable optical add/drop multiplexer (ROADM) networks using machine learning. In one example, a method includes defining a feature set for a proposed path through a wavelength division multiplexing network, wherein the proposed path traverses at least one link in the network, and wherein the at least one link connects a pair of reconfigurable optical add/drop multiplexers, predicting an optical performance of the proposed path, wherein the predicting employs a machine learning model that takes the feature set as an input and outputs a metric that quantifies predicted optical performance, and determining whether to deploy a new wavelength on the proposed path based on the predicted optical performance of the proposed path.

    Traffic matrix prediction and fast reroute path computation in packet networks
    13.
    发明授权
    Traffic matrix prediction and fast reroute path computation in packet networks 有权

    公开(公告)号：US10581736B1

    公开(公告)日：2020-03-03

    申请号：US16189786

    申请日：2018-11-13

    Applicant: AT&T Intellectual Property I, L.P.

    Inventor： Gagan Choudhury ,  Gaurav Thakur ,  Simon Tse ,  Kathleen Meier-Hellstern

    IPC: H04B10/00 ,  H04L12/721 ,  H04L12/715 ,  H04L12/26 ,  H04Q11/00 ,  H04L12/723

    Abstract: A processing system including at least one processor may obtain traffic measurements for end-to-end paths in a telecommunication network, calculate traffic estimates for the end-to-end paths in future time periods based on the traffic measurements in accordance with at least one machine learning model, calculate traffic estimates for primary paths in the telecommunication network based upon the traffic estimates for the end-to-end paths, compute a backup path configuration for a primary path of the telecommunication network for the future time periods based upon the traffic estimates for the primary paths in the future time periods, detect a change in the backup path configuration for the primary path in a future time period based upon the computing, and adjust a backup path in accordance with the backup path configuration when the change in the backup path configuration is detected.

    MACHINE LEARNING TECHNIQUES FOR SELECTING PATHS IN MULTI-VENDOR RECONFIGURABLE OPTICAL ADD/DROP MULTIPLEXER NETWORKS
    15.
    发明申请
    MACHINE LEARNING TECHNIQUES FOR SELECTING PATHS IN MULTI-VENDOR RECONFIGURABLE OPTICAL ADD/DROP MULTIPLEXER NETWORKS 有权

    公开(公告)号：US20210306086A1

    公开(公告)日：2021-09-30

    申请号：US17347383

    申请日：2021-06-14

    Applicant: AT&T Intellectual Property I, L.P.

    Inventor： Martin Birk ,  David Frederick Lynch ,  Gaurav Thakur ,  Simon Tse

    IPC: H04J14/02 ,  H04L12/751 ,  H04B10/079

    Abstract: Devices, computer-readable media and methods are disclosed for selecting paths in reconfigurable optical add/drop multiplexer (ROADM) networks using machine learning. In one example, a method includes defining a feature set for a proposed path through a wavelength division multiplexing network, wherein the proposed path traverses at least one link in the network, and wherein the at least one link connects a pair of reconfigurable optical add/drop multiplexers, predicting an optical performance of the proposed path, wherein the predicting employs a machine learning model that takes the feature set as an input and outputs a metric that quantifies predicted optical performance, and determining whether to deploy a new wavelength on the proposed path based on the predicted optical performance of the proposed path.

    Cell site placement system
    16.
    发明授权
    Cell site placement system 有权

    公开(公告)号：US11082862B1

    公开(公告)日：2021-08-03

    申请号：US16842071

    申请日：2020-04-07

    Applicant: AT&T Intellectual Property I, L.P.

    Inventor： Gaurav Thakur ,  Danielle Liu ,  Arun Jotshi ,  Ashwin Sridharan ,  Kathleen Meier-Hellstern

    IPC: H04W16/18 ,  G06K9/00 ,  G06T7/30 ,  H04W24/08

    Abstract: A device includes a processor and a memory. The processor effectuates operations including generating a plurality of tiles for a designated region and classifying each of the plurality of tiles based on whether each tile is associated with a deployment zone. The operations further including clustering locations associated with one or more tiles of the plurality of tiles, wherein the one or more tiles are classified as being associated with the deployment zone, wherein the clustering generates at least one vertex. The operations further including forming a polygon based on the at least one vertex. The operations further including providing a map of the designated region including the polygon at a location in the map associated with the deployment zone, wherein the deployment zone is one or more areas of service within the designated region that are prioritized for new or additional services, or infrastructure based on design criteria.

    Creating and using network coverage models
    17.
    发明授权
    Creating and using network coverage models 有权

    公开(公告)号：US10959109B1

    公开(公告)日：2021-03-23

    申请号：US16802867

    申请日：2020-02-27

    Applicant: AT&T Intellectual Property I, L.P.

    Inventor： Danielle Liu ,  Arun Jotshi ,  Gaurav Thakur ,  Velin Kounev ,  Fei Wu ,  Yaron Kanza ,  Divith Aruni Babu

    IPC: H04W16/18 ,  G06N3/04 ,  G06N3/08 ,  G06K9/00 ,  H04W64/00

    Abstract: Concepts and technologies are disclosed herein for creating and using network coverage models. A request for a predicted coverage model that represents a first signal propagation in a first portion of a network that covers a first area associated with a first geographic location can be received. An aerial image that depicts the first area can be obtained. The aerial image can be provided to an existing coverage model. The existing coverage model can include a neural network, and the existing coverage model can be based on a second signal propagation in a second portion of the network that covers a second area associated with a second location. The predicted coverage model for the first area can be obtained from the existing coverage model.

    Traffic matrix prediction and fast reroute path computation in packet networks
    18.
    发明授权
    Traffic matrix prediction and fast reroute path computation in packet networks 有权

    公开(公告)号：US10805214B2

    公开(公告)日：2020-10-13

    申请号：US16806837

    申请日：2020-03-02

    Applicant: AT&T Intellectual Property I, L.P.

    Inventor： Gagan Choudhury ,  Gaurav Thakur ,  Simon Tse ,  Kathleen Meier-Hellstern

    IPC: H04B10/00 ,  H04L12/721 ,  H04L12/715 ,  H04L12/26 ,  H04Q11/00 ,  H04L12/723

    Abstract: A processing system including at least one processor may obtain traffic measurements for end-to-end paths in a telecommunication network, calculate traffic estimates for the end-to-end paths in future time periods based on the traffic measurements in accordance with at least one machine learning model, calculate traffic estimates for primary paths in the telecommunication network based upon the traffic estimates for the end-to-end paths, compute a backup path configuration for a primary path of the telecommunication network for the future time periods based upon the traffic estimates for the primary paths in the future time periods, detect a change in the backup path configuration for the primary path in a future time period based upon the computing, and adjust a backup path in accordance with the backup path configuration when the change in the backup path configuration is detected.

Patent Agency Ranking