中科微点-全球专利检索

  1. Login
  2. Sign Up

Search Results

35 records (took 0.02 seconds)

    MULTI-OBJECTIVE NEURAL ARCHITECTURE SEARCH FRAMEWORK
    11.
    发明公开
    MULTI-OBJECTIVE NEURAL ARCHITECTURE SEARCH FRAMEWORK 审中-公开

    公开(公告)号：US20240176986A1

    公开(公告)日：2024-05-30

    申请号：US18114658

    申请日：2023-02-27

    Applicant: Samsung Electronics Co., Ltd.

    Inventor： Mostafa EL-KHAMY ,  Minsu CHO ,  Kee-Bong SONG

    IPC: G06N3/0442 ,  G06N3/086 ,  G06N3/092

    CPC classification number: G06N3/0442 ,  G06N3/086 ,  G06N3/092

    Abstract: A system and a method are disclosed for performing a neural architecture search. The method includes sampling a discrete network search space a first time, determining a differential architecture network sampled from a super-network using continuous relaxation of the discrete network search space over operators in the super-network, calculating a reward based on a proxy accuracy or a proxy complexity of the differential architecture network, updating a distribution of the discrete network search space based on the reward, and determining an updated differential architecture network based on the reward.

    DISTRIBUTED CLOSED-LOOP POWER CONTROL WITH VGA GAIN UPDATE
    12.
    发明公开
    DISTRIBUTED CLOSED-LOOP POWER CONTROL WITH VGA GAIN UPDATE 审中-公开

    公开(公告)号：US20230353195A1

    公开(公告)日：2023-11-02

    申请号：US17731251

    申请日：2022-04-27

    Applicant: Samsung Electronics Co., Ltd.

    Inventor： Yanru TANG ,  Pranav DAYAL ,  Hou-Shin CHEN ,  Pritesh VORA ,  Tienyu CHANG ,  Siuchuang Ivan LU ,  Kee-Bong SONG

    IPC: H04B7/04 ,  H03F3/24

    CPC classification number: H04B7/04 ,  H03F3/245 ,  H04B2001/0416

    Abstract: A closed-loop power control (CLPC) system is disclosed that includes a first signal path for a first polarization and a second signal path for a second polarization. The first signal path includes a first power amplifier, a first output power detector configured to detect a first output power level of the first power amplifier, and a first processor configured to determine a first analog gain for a first controller and a first gain for a first digital-to-analog converter based on a first accumulated error between the first output power level and a target Effective Isotropically Radiated Power. A second processor is configured to set a first variable gain of a first variable gain amplifier coupled to an input of the first power amplifier. The CLPC can be configured to control the gain of the first signal path separately or as one signal path.

    HARDWARE ALLOCATION IN RFIC BASED ON MACHINE LEARNING
    18.
    发明公开
    HARDWARE ALLOCATION IN RFIC BASED ON MACHINE LEARNING 审中-公开

    公开(公告)号：US20230297803A1

    公开(公告)日：2023-09-21

    申请号：US17834845

    申请日：2022-06-07

    Applicant: Samsung Electronics Co., Ltd.

    Inventor： Shailesh CHAUDHARI ,  Hyukjoon KWON ,  Kee-Bong SONG ,  Dongwoon BAI

    IPC: G06K19/07 ,  G06N3/08

    CPC classification number: G06K19/0723 ,  G06N3/08

    Abstract: A system and method for configuring an RF network based on machine learning. In some embodiments, the method includes: receiving, by a first neural network, a first state and a first state transition, the first state including: one or more identifiers for available active ports, and a set of available connections between two or more circuit elements, each of the circuit elements being one of: (1) a first circuit type, (2) a second circuit type that operatively connects a circuit element of the first circuit type to one of the available active ports, and (3) the available active ports; and generating, by the first neural network, a first estimated quality value, for the first state transition.

Patent Agency Ranking