中科微点-全球专利检索

  1. Login
  2. Sign Up

Search Results

17 records (took 0.016 seconds)

    METHOD AND APPARATUS FOR LATTICE-BASED NON-LINEAR PRECODING FOR NON-ORTHOGONAL MULTIPLE ACCESS
    1.
    发明申请
    METHOD AND APPARATUS FOR LATTICE-BASED NON-LINEAR PRECODING FOR NON-ORTHOGONAL MULTIPLE ACCESS 审中-公开
    用于非正交多路访问的基于网格的非线性预编码的方法和装置

    公开(公告)号：WO2018010807A1

    公开(公告)日：2018-01-18

    申请号：PCT/EP2016/066831

    申请日：2016-07-14

    Applicant: HUAWEI TECHNOLOGIES CO., LTD. ,  BENAMMAR, Meryem ,  ESTELLA AGUERRI, Inaki ,  ZAIDI, Abdellatif ,  BELFIORE, Jean-Claude

    Inventor： BENAMMAR, Meryem ,  ESTELLA AGUERRI, Inaki ,  ZAIDI, Abdellatif ,  BELFIORE, Jean-Claude

    IPC: H04W52/34 ,  H04W52/42

    CPC classification number: H04W52/346 ,  H04W52/143 ,  H04W52/42

    Abstract: The present invention relates to a wireless communication system for downlink transmitting, through a plurality of orthogonal carriers and according to a non-orthogonal multiple access transmission scheme in a single-input single-output configuration, a plurality of messages towards, respectively, a plurality of users. At the transmitter side, a power allocation strategy subjected to a total power budget constraint is implemented to meet a prescribed fairness constraint. The interference caused to each other by the users served at identical resources is efficiently mitigated through an intra-carrier lattice-based non-linear precoding process operating sequentially and taking account of the power allocation strategy, the users' ordering and the channel state information of each channel linking the respective carrier to each user. At the receiver side, the plurality of messages is respectively recovered by each user thanks to a respective simple single-user decoder, which decodes only its intended signal without requiring any successive interference cancellation procedure.

    Abstract translation: 本发明涉及一种无线通信系统，用于通过多个正交载波并根据单输入单输出配置中的非正交多址接入传输方案来下行传输多个 的消息分别朝向多个用户。 在发射机侧，实施受总功率预算约束的功率分配策略以满足规定的公平性约束。 通过顺序操作并且考虑功率分配策略，用户的排序以及信道状态信息来有效地缓解服务于相同资源的用户彼此造成的干扰 每个信道将相应的载波链接到每个用户。 在接收机侧，由于各个简单的单用户解码器，每个用户分别恢复多个消息，该单用户解码器仅解码其预期信号，而不需要任何连续的干扰消除过程。

    DEVICE AND METHOD FOR A BOOLEAN NEURAL NETWORK
    2.
    发明申请
    DEVICE AND METHOD FOR A BOOLEAN NEURAL NETWORK 审中-公开

    公开(公告)号：WO2022100827A1

    公开(公告)日：2022-05-19

    申请号：PCT/EP2020/081814

    申请日：2020-11-12

    Applicant: HUAWEI TECHNOLOGIES CO., LTD. ,  NGUYEN, Van Minh

    Inventor： BELFIORE, Jean-Claude ,  EL JABRI, Mejed

    IPC: G06N3/04 ,  G06N3/08

    Abstract: Various embodiments relate to neural networks. Neurons may be configured with binary weights associated with respective binary inputs. A Boolean function may be evaluated for the binary weights and respective binary inputs. A pre-activation may be determined based on a sum of binary outputs of the Boolean function. A binary output of a neuron may be determined based on a threshold for the pre-activation. Training a neural network may comprise receiving binary backpropagation signal(s) from a downstream layer. The binary backpropagation signal(s) may indicate a tendency of a loss function with respect to variation of the binary output. A tendency of the Boolean function with respect to inversion of an ith binary weight may be determined. Training may comprise determining whether to invert the ith binary weight based on the tendency of the Boolean function and the tendency of the loss function. Devices, methods, and computer programs are disclosed.

    INTEGRATED TRAINING AND INFERENCE OF NEURONS AND NEURAL NETWORKS
    3.
    发明申请
    INTEGRATED TRAINING AND INFERENCE OF NEURONS AND NEURAL NETWORKS 审中-公开

    公开(公告)号：WO2022083849A1

    公开(公告)日：2022-04-28

    申请号：PCT/EP2020/079413

    申请日：2020-10-20

    Applicant: HUAWEI TECHNOLOGIES CO., LTD. ,  LIAKOPOULOS, Nikolaos

    Inventor： ARVANITAKIS, George ,  TSILIMANTOS, Dimitrios ,  BELFIORE, Jean-Claude

    IPC: G06N3/04 ,  G06N3/08

    Abstract: A computing device comprising or configured to implement a neural network for operating on a received input by executing a learned process on said input to produce an output, the neural network being implemented by a main module configured for both training the neural network and inference of a machine learning model. The main module comprising a projection module and a coding module. The projection module is configured to project an input vector from an initial vector space into a codeword space based on a machine learning model stored for the projection module. The coding module is configured to evaluate the machine learning model stored for the projection module and form evaluation data representing the result of that evaluation. The projection module is also configured to adapt the machine learning model stored for the projection module in dependence on the evaluation data.

    PUNCTURING AND SHORTENING OF POLAR CODES
    4.
    发明申请
    PUNCTURING AND SHORTENING OF POLAR CODES 审中-公开
    极化代码的调整和缩​​短

    公开(公告)号：WO2017194133A1

    公开(公告)日：2017-11-16

    申请号：PCT/EP2016/060730

    申请日：2016-05-12

    Applicant: HUAWEI TECHNOLOGIES CO., LTD. ,  BIOGLIO, Valerio ,  GABRY, Frederic ,  LAND, Ingmar ,  BELFIORE, Jean-Claude

    Inventor： BIOGLIO, Valerio ,  GABRY, Frederic ,  LAND, Ingmar ,  BELFIORE, Jean-Claude

    IPC: H03M13/00 ,  H03M13/13

    Abstract: Encoder (400) for encoding K information bits into a code word of length N´ on the basis of a polar code of length N, wherein N is a power of 2 and greater than or equal to N´. The encoder (400) comprises a memory (403) storing a plurality of bit indices, wherein the plurality of bit indices comprise a set of N frozen bit indices associated with the polar code of length N, a set of N/2 puncturing bit indices and/or a set of N/2 shortening bit indices and a processor (401) configured to retrieve at least a subset of the plurality of bit indices from the memory (403), to encode the K information bits using the polar code of length N for obtaining encoded data of length N and to reduce the number of bits of the encoded data to the length N´ for obtaining the code word of length N´.

    Abstract translation: 用于将K个信息比特编码为长度为N的码字的编码器（400） 基于长度为N的极性码，其中N是2的幂并且大于或等于N. 编码器（400）包括存储多个比特索引的存储器（403），其中多个比特索引包括与长度为N的极性码相关联的一组N个冻结比特索引，一组N / 2个打孔比特索引 和/或一组N / 2个缩短比特索引，以及处理器（401），被配置为从存储器（403）检索多个比特索引的至少一个子集，以使用长度极性码对K个信息比特进行编码 N，用于获得长度为N的编码数据并将编码数据的比特数减少到长度N; 用于获取长度为N的码字。

    A METHOD TO GENERATE ORDERED SEQUENCE FOR POLAR CODES
    6.
    发明申请
    A METHOD TO GENERATE ORDERED SEQUENCE FOR POLAR CODES 审中-公开

    公开(公告)号：WO2018153263A1

    公开(公告)日：2018-08-30

    申请号：PCT/CN2018/075605

    申请日：2018-02-07

    Applicant: HUAWEI TECHNOLOGIES CO., LTD.

    Inventor： BELFIORE, Jean-Claude ,  GE, Yiqun ,  HE, Gaoning ,  ZHANG, Ran ,  LAND, Ingmar ,  SHI, Wuxian ,  TONG, Wen

    IPC: H03M13/13

    Abstract: A number K of N sub-channels that are defined by a code and that have associated reliabilities for input bits at N input bit positions, are to be selected to carry bits that are to be encoded. A localization area that includes multiple sub-channels and is located below fewer than K of the N sub-channels in a partial order of the N sub-channels is determined based on one or more coding parameters. The fewer than K sub-channels of the N sub-channels above the localization area in the partial order are selected, and a number of sub-channels from those in the localization area are also selected. The selected fewer than K sub-channels and the number of sub-channels selected from those in the localization area together include K sub-channels to carry the bits that are to be encoded.

    DEVICES AND METHODS FOR CSI-ASSISTED BEAMFORMING USING A LEARNING NETWORK ENTITY
    9.
    发明申请
    DEVICES AND METHODS FOR CSI-ASSISTED BEAMFORMING USING A LEARNING NETWORK ENTITY 审中-公开

    公开(公告)号：WO2020043284A1

    公开(公告)日：2020-03-05

    申请号：PCT/EP2018/073265

    申请日：2018-08-29

    Applicant: HUAWEI TECHNOLOGIES CO., LTD. ,  GUILLAUD, Maxime

    Inventor： GUILLAUD, Maxime ,  HE, Gaoning ,  GARCIA ORDONEZ, Luis ,  FERRAND, Paul ,  DECURNINGE, Alexis ,  BELFIORE, Jean-Claude ,  YANG, Ganghua

    IPC: H04B7/06 ,  H04B7/08 ,  H04W74/00

    Abstract: The invention relates to a network entity (120) for a cellular communication network (100).The cellular communication network (100) comprises a radio base station (101), a millimeter wave base station (101) and a mobile terminal (111) configured to communicate with the radio base station (101) via a radio communication channel and to communicate with the millimeter wave base station (101) using a millimeter beam defining a beam direction of a plurality of beam directions (111a-d). The network entity (120) comprises: a communication interface (123) configured to collect information about the state of the radio communication channel and information about the beam direction for a plurality of different positions of the mobile terminal (111) and/or for a plurality of different points in time; and a processor (121) implementing a neural network. The neural network is configured to: in a training phase, determine a correlation between the information about the state of the radio communication channel and the information about the beam direction for the plurality of different positions of the mobile terminal (111) and/or for the plurality of different points in time; and in an exploitation phase, predict one or more target beam directions on the basis of the correlation and information about a current state of the radio communication channel.

    CONSTRUCTION OF A POLAR CODE BASED ON A DISTANCE CRITERION AND A RELIABILITY CRITERION, IN PARTICULAR OF A MULTI-KERNEL POLAR CODE
    10.
    发明申请
    CONSTRUCTION OF A POLAR CODE BASED ON A DISTANCE CRITERION AND A RELIABILITY CRITERION, IN PARTICULAR OF A MULTI-KERNEL POLAR CODE 审中-公开

    公开(公告)号：WO2019020182A1

    公开(公告)日：2019-01-31

    申请号：PCT/EP2017/068930

    申请日：2017-07-26

    Applicant: HUAWEI TECHNOLOGIES CO., LTD. ,  BIOGLIO, Valerio ,  LAND, Ingmar ,  BELFIORE, Jean-Claude ,  GABRY, Frederic

    Inventor： BIOGLIO, Valerio ,  LAND, Ingmar ,  BELFIORE, Jean-Claude ,  GABRY, Frederic

    IPC: H03M13/13

    Abstract: The present invention relates to a device (102b, 104b) for generating a polar code ϰ N of length N and dimension K on the basis of a transformation matrix G N of size N x N , wherein the transformation matrix G N is based on a first matrix G N r of size N r x N r , and on a second matrix G Nd of size N d x N d , wherein N = N r ⋅ N d , and wherein the polar code ϰ N is given by ϰ N = υ N ⋅ G N , wherein υ N = (υ 0 ,..., υ N -1 )is a vector of size N , υ i , i = 0,... N −1, corresponding to an information bit if i ε I , I being a set of K information bit indices, and υ i = 0, if i ε F , F being a set of N − K frozen bit indices. The device (102b, 104b) comprises a processor (102c, 104c) configured to generate a reliability vector v GNr = [v 1 ,... v Nr ], wherein v i represents a reliability of an i -th input bit of a code generated by the first matrix G Nr , generate a distance spectrum vector d GNd = [d 1 ,..., d Nd ] of a code generated by the second matrix G Nd , wherein d j represents a minimum distance of the code generated by the second matrix G Nd of dimension j , determine the set of K information bit indices I on the basis of the reliability vector v GNr and of the distance spectrum vector d GN , and generate the polar code c N on the basis of the set of K information bit indices I .

Patent Agency Ranking