GRADIENT ACCUMULATION FOR FEDERATED LEARNING
    132.
    发明公开

    公开(公告)号:US20230232377A1

    公开(公告)日:2023-07-20

    申请号:US17648117

    申请日:2022-01-14

    CPC classification number: H04W72/044 G06N20/00 H04W24/02

    Abstract: A UE may identify, in each round other than an initial round, a first plurality of local model update elements of a present round. The first plurality of local model update elements of the present round may be associated with an updated local machine learning model. The UE may transmit to a base station, in each round other than the initial round, over a multiple access channel via analog signaling, a second plurality of local model update elements of the present round based on a third plurality of local model update elements of the present round. The third plurality of local model update elements of the present round may correspond to a sum of the first plurality of local model update elements of the present round and a local model update error of a previous round immediately before the present round. The analog signaling may be associated with OTA aggregation.

    GAIN SCALING OF INPUT TO NEURAL NETWORK FOR END-TO-END LEARNING IN WIRELESS COMMUNICATION SYSTEM

    公开(公告)号:US20230114870A1

    公开(公告)日:2023-04-13

    申请号:US17498651

    申请日:2021-10-11

    Abstract: A method of wireless communication by a user equipment (UE) includes receiving different sets of parameters from different sources as input to a receiver neural network. The method also includes receiving, from a base station, a set of target long-term energy values associated with the receiver neural network. The method further includes calculating a scaling factor for each of the different sets of parameters based on the set of target long-term energy values, and separately scaling each of the different sets of parameters based on the scaling factor calculated for that set in order to generate multiple sets of scaled parameters. The method still further includes transmitting the multiple sets of scaled parameters to the receiver neural network.

    DETERMINING OVERLAY CODES FOR TRANSMISSION OF REFERENCE SIGNALS

    公开(公告)号:US20220216938A1

    公开(公告)日:2022-07-07

    申请号:US17142172

    申请日:2021-01-05

    Abstract: Disclosed are techniques for determining tone patterns and associated overlay codes for transmission of reference signals. A tone pattern (e.g., with each tone pattern occupying a resource element in a resource block) can be determined for a reference signal for use in wireless communications between a receiving device and a transmitting device. The tone pattern can include an irregular combination of resource elements in one or more resource blocks for the reference signal. The resource elements can be shared by a plurality of antennas for communication of one or more reference signals between at least the user equipment and the base station. An overlay code may be determined for the tone pattern. Information associated with the transmission of the reference signal using the tone pattern and the overlay code may be communicated by the receiving device.

Patent Agency Ranking