中科微点-全球专利检索

  1. Login
  2. Sign Up

Search Results

17459 records (took 0.074 seconds)

    DELAY ENCODED VECTOR SYMBOLIC RADIO MULTIPLE ACCESS
    1.
    发明申请
    DELAY ENCODED VECTOR SYMBOLIC RADIO MULTIPLE ACCESS 审中-公开

    公开(公告)号：WO2023091061A1

    公开(公告)日：2023-05-25

    申请号：PCT/SE2022/051050

    申请日：2022-11-10

    Applicant: TELEFONAKTIEBOLAGET LM ERICSSON (PUBL)

    Inventor： VERES, András ,  RÁCZ, András ,  HÁGA, Péter ,  KENESI, Zsolt ,  BORSOS, Tamas

    IPC: G06N3/049 ,  G06N3/063 ,  H04B1/00 ,  H04B7/0413 ,  H04L5/00 ,  H04L27/26 ,  H04L27/20 ,  H04W16/00 ,  H04W80/00 ,  H04W84/00

    Abstract: A method by a neuromorphic device in a wireless communication network which communicates using radio frames and carrier frequencies. The method includes obtaining a high dimensional (HD) vector containing symbols. At least some symbols have a value indicating a pattern of firing events for associated one or more neurons of a neural network (NN). For each symbol in the HD vector having a nonzero value, selecting a subframe of a radio frame and/or a carrier frequency among a set of carrier frequencies, based on a defined mapping between subframes of the radio frame and/or carrier frequencies of the set and the locations of symbols in the HD vector, determining a time offset relative to the selected subframe and/or a frequency offset relative to the selected carrier frequency, based on the value of the symbol, and transmitting an impulse at the determined time offset relative to the selected subframe and/or at the determined frequency offset relative to the selected carrier frequency. (Figure 7)

    AN AUTOENCODER FOR DATA COMPRESSION
    4.
    发明申请
    AN AUTOENCODER FOR DATA COMPRESSION 审中-公开

    公开(公告)号：WO2023088562A1

    公开(公告)日：2023-05-25

    申请号：PCT/EP2021/082306

    申请日：2021-11-19

    Applicant: TELEFONAKTIEBOLAGET LM ERICSSON (PUBL)

    Inventor： LI, Yun

    IPC: G06N3/04 ,  G06N3/08 ,  H04N19/00 ,  G06N7/00

    Abstract: A computer-implemented method of training an autoencoder for data compression is provided. The method comprises inputting training data into a neural network to generate a plurality of first latent variables, using a probability estimator to obtain probability mass functions for the plurality of latent variables, and determining a rate measure based on the probability mass functions and a plurality of coded latent variables. The plurality of coded latent variables is based on one-hot encoding of plurality of quantized latent variables obtained by quantizing the plurality of latent variables. A gradient of the rate measure is determined based on the probability mass functions and a plurality of approximately coded latent variables obtained by applying a differentiable function to the plurality of latent variables to approximate one-hot encoding. One or more parameters of the autoencoder is updated based on the rate measure and the gradient of the rate measure.

    SERVER AND AGENT FOR REPORTING OF COMPUTATIONAL RESULTS DURING AN ITERATIVE LEARNING PROCESS
    5.
    发明申请
    SERVER AND AGENT FOR REPORTING OF COMPUTATIONAL RESULTS DURING AN ITERATIVE LEARNING PROCESS 审中-公开

    公开(公告)号：WO2023088533A1

    公开(公告)日：2023-05-25

    申请号：PCT/EP2021/081773

    申请日：2021-11-16

    Applicant: TELEFONAKTIEBOLAGET LM ERICSSON (PUBL)

    Inventor： MOOSAVI, Reza ,  RYDÉN, Henrik ,  LARSSON, Erik, G.

    IPC: G06N20/20 ,  H04B1/707 ,  H04L41/16 ,  H04W52/14 ,  H04W52/22 ,  H04W52/24 ,  H04W52/38

    Abstract: There is provided mechanisms for performing an iterative learning process with agent entities. A method is performed by a server entity. The method comprises selecting a linear mapping to be used by the agent entities when reporting computational results of a computational task to the server entity. The linear mapping has an inverse. The linear mapping is selected as a function of an estimated interference level for a radio propagation channel over which the agent entities are to report the computational results of the computational task to the server entity. The method comprises configuring the agent entities with the computational task. The method comprises performing the iterative learning process with the agent entities until a termination criterion is met. The server entity as part of performing the iterative learning process applies the inverse of the linear mapping to a sum of the computational results received from the agent entities.

    METHOD AND APPARATUS FOR ANTENNA CALIBRATION CONTROL
    6.
    发明申请
    METHOD AND APPARATUS FOR ANTENNA CALIBRATION CONTROL 审中-公开

    公开(公告)号：WO2023087170A1

    公开(公告)日：2023-05-25

    申请号：PCT/CN2021/131145

    申请日：2021-11-17

    Applicant: TELEFONAKTIEBOLAGET LM ERICSSON (PUBL) ,  ZHANG, Hao

    Inventor： NILSSON, Magnus ,  BRAUN, Christian ,  ZHANG, Hao

    IPC: H04B17/14

    Abstract: A method is disclosed which is for controlling antenna calibration of an antenna system residing in a wireless communication network. The antenna calibration is for compensation of impairments associated with operation of the antenna system. The method comprises performing over-the-air measurements of the antenna calibration during unused symbols defined by a predetermined pattern of unused symbols, wherein the predetermined pattern of unused symbols comprises unused symbols directly preceding synchronization symbols and/or unused symbols directly following synchronization symbols. For example, the unused symbols of the predetermined pattern of unused symbols may be symbols of slots that comprise one or more synchronization signal blocks (SSB). In some embodiments, the predetermined pattern comprises sequences of consecutive unused symbols. Each sequence of consecutive unused symbols which has a length exceeding a measurement length threshold may comprise one or more measurement opportunities, and each measurement may be performed during a respective measurement opportunity.

    PROTECTING A MODEL AGAINST AN ADVERSARY
    7.
    发明申请
    PROTECTING A MODEL AGAINST AN ADVERSARY 审中-公开

    公开(公告)号：WO2023085984A1

    公开(公告)日：2023-05-19

    申请号：PCT/SE2021/051126

    申请日：2021-11-10

    Applicant: TELEFONAKTIEBOLAGET LM ERICSSON (PUBL)

    Inventor： KIM, Seonghyun ,  LILJENSTAM, Michael ,  STERNBY, Jakob ,  BARRIGA, Luis

    IPC: G06F21/55 ,  G06F21/60 ,  G06N5/02 ,  G06N20/00

    Abstract: A method for use in protecting a first machine learning model that is queryable over an application programming interface, API, against an adversary querying the first machine learning model through the API in order to build up a database of query-response pairs. The method comprises: identifying (202) a user of the API as a potential adversary. In response to a query from the potential adversary, through the API, then method comprises providing (204) a response from a second machine learning model instead of the first machine learning model, wherein the first machine learning model has been trained on a first dataset and wherein the second machine learning model has been trained on a second dataset that is different to the first dataset.

    GENERATING A SECURITY METRIC USING A MACHINE LEARNING MODEL
    9.
    发明申请
    GENERATING A SECURITY METRIC USING A MACHINE LEARNING MODEL 审中-公开

    公开(公告)号：WO2023084292A1

    公开(公告)日：2023-05-19

    申请号：PCT/IB2021/060489

    申请日：2021-11-12

    Applicant: TELEFONAKTIEBOLAGET LM ERICSSON (PUBL)

    Inventor： FARRAHI MOGHADDAM, Fereydoun

    IPC: H04L9/40 ,  G06N20/00

    Abstract: A method (400) for generating a security metric for a target system is disclosed. The security metric provides a quantitative representation of the security of the target system, which representation is non-specific to any given component of the system or point of attack. The method involves generation of an analogical model of the target system (410), and mapping of the analogical model to a directed graph of the target system (430). A distance metric is then calculated (440) between the directed graph of the target system and directed graphs of a plurality of reference systems. The method further comprises inputting a tensor comprising the calculated distance metrics to a trained ML model (450), wherein the ML model has been trained using a training data set that is based on the same plurality of reference systems (450a) and wherein the ML model is operable to process the input tensor and to generate an output comprising a value of the security metric for the target system (450b). Also disclosed is a method (300) for training an ML model to generate a security metric.

    GENERATING AN AUTHENTICATION TOKEN
    10.
    发明申请
    GENERATING AN AUTHENTICATION TOKEN 审中-公开

    公开(公告)号：WO2023083691A1

    公开(公告)日：2023-05-19

    申请号：PCT/EP2022/080688

    申请日：2022-11-03

    Applicant: TELEFONAKTIEBOLAGET LM ERICSSON (PUBL)

    Inventor： ORSINO, Antonino ,  NAKARMI, Prajwol Kumar

    IPC: H04W12/06 ,  H04W76/19

    Abstract: Embodiments of the present disclosure includes methods for generating an authentication token. One of the methods comprising obtaining a first request message that comprises: a first field containing a first value, a second field containing a second value, and a third field containing a pre-defined token value; and using the first request message to generate an authentication token.

Patent Agency Ranking