Abstract: A device for implementing a tensor-train decomposition operation for a respective convolutional layer of a convolutional neural network (CNN) is provided. The device is configured to receive input data comprising a first number of channels, and perform a 1×1 convolution on the input data to obtain a plurality of data groups. The plurality of data groups comprises a second number of channels. The device is further configured to perform a group convolution on the plurality of data groups to obtain intermediate data comprising a third number of channels, and perform a 1×1 convolution on the intermediate data to obtain output data comprising a fourth number of channels.

Abstract: Embodiments of this application disclose provides a low density parity check (LDPC) channel encoding method for use in a wireless communications system. A communication device encodes an input bit sequence by using a LDPC matrix, to obtain an encoded bit sequence for transmission. The LDPC matrix is obtained based on a lifting factor Z and a base matrix. Embodiments of the application provide eight particular designs of the base matrix. The encoding method provided in the embodiments of the application can be used in various communications systems including the fifth generation (5G) telecommunication systems, and can support various encoding requirements for information bit sequences with different code lengths.

Abstract: A low density parity check (LDPC) channel encoding method is used in a wireless communications system. A communication device encodes an input bit sequence by using an LDPC matrix, to obtain an encoded bit sequence for transmission. The LDPC matrix is obtained based on a lifting factor Z and a base matrix. The base matrix may be one of eight exemplary designs. The encoding method can be used in various communications systems including fifth generation (5G) telecommunication systems, and can support various encoding requirements for information bit sequences with different code lengths.

Abstract: The present disclosure relates to processing input data by a neural network. Methods and apparatuses of some embodiments process the input data by at least one layer of the neural network and obtain thereby a feature tensor. Then, the distribution of the obtained feature tensors estimated. Another distribution is obtained. Such other distribution may be a distribution of another input data, or a distribution obtained by combining a plurality of distributions obtained for respective plurality of some input data. Then a distance value indicative of a distance between the two distributions is calculated and based thereon, a characteristic of the input data is determined. The characteristic may be pertinence to a certain class of data or a detection of out-of-distribution data or determination of reliability of a class determination or the like.

Abstract: Embodiments of this application disclose provides a low density parity check (LDPC) channel encoding method for use in a wireless communications system. A communication device encodes an input bit sequence by using a LDPC matrix, to obtain an encoded bit sequence for transmission. The LDPC matrix is obtained based on a lifting factor Z and a base matrix. Embodiments of the application provide eight particular designs of the base matrix. The encoding method provided in the embodiments of the application can be used in various communications systems including the fifth generation (5G) telecommunication systems, and can support various encoding requirements for information bit sequences with different code lengths.

Abstract: A method for generating a code, a method for encoding and decoding data, and an encoder and a decoder performing the encoding and decoding are disclosed. In an embodiment, a method for lifting a child code from a base code for encoding and decoding data includes determining a single combination of a circulant size, a lifting function, and a labelled base matrix PCM according to an information length and a code rate using data stored in a lifting table. The lifting table was defined at a code generation stage. The method also includes calculating a plurality of shifts for the child code. Each shift is calculated by applying the lifting function to the labelled base matrix PCM with a defined index using the circulant size and using the derived child PCM to encode or decode data.

Abstract: A low density parity check (LDPC) channel encoding method is used in a wireless communications system. A communication device encodes an input bit sequence by using an LDPC matrix, to obtain an encoded bit sequence for transmission. The LDPC matrix is obtained based on a lifting factor Z and a base matrix. The base matrix may be one of eight exemplary designs. The encoding method can be used in various communications systems including fifth generation (5G) telecommunication systems, and can support various encoding requirements for information bit sequences with different code lengths.

Abstract: A low density parity check (LDPC) channel encoding method is used in a wireless communications system. A communication device encodes an input bit sequence by using an LDPC matrix, to obtain an encoded bit sequence for transmission. The LDPC matrix is obtained based on a lifting factor Z and a base matrix. The base matrix may be one of eight exemplary designs. The encoding method can be used in various communications systems including fifth generation (5G) telecommunication systems, and can support various encoding requirements for information bit sequences with different code lengths.

Abstract: Embodiments of this application disclose provides a low density parity check (LDPC) channel encoding method for use in a wireless communications system. A communication device encodes an input bit sequence by using a LDPC matrix, to obtain an encoded bit sequence for transmission. The LDPC matrix is obtained based on a lifting factor Z and a base matrix. Embodiments of the application provide eight particular designs of the base matrix. The encoding method provided in the embodiments of the application can be used in various communications systems including the fifth generation (5G) telecommunication systems, and can support various encoding requirements for information bit sequences with different code lengths.

Abstract: A low density parity check (LDPC) channel encoding method is used in a wireless communications system. A communication device encodes an input bit sequence by using an LDPC matrix, to obtain an encoded bit sequence for transmission. The LDPC matrix is obtained based on a lifting factor Z and a base matrix. The base matrix may be one of eight exemplary designs. The encoding method can be used in various communications systems including fifth generation (5G) telecommunication systems, and can support various encoding requirements for information bit sequences with different code lengths.

Abstract: A method for generating a code, a method for encoding and decoding data, and an encoder and a decoder performing the encoding and decoding are disclosed. In an embodiment, a method for lifting a child code from a base code for encoding and decoding data includes determining a single combination of a circulant size, a lifting function, and a labelled base matrix PCM according to an information length and a code rate using data stored in a lifting table. The lifting table was defined at a code generation stage. The method also includes calculating a plurality of shifts for the child code. Each shift is calculated by applying the lifting function to the labelled base matrix PCM with a defined index using the circulant size and using the derived child PCM to encode or decode data.

Abstract: Embodiments of this application disclose provides a low density parity check (LDPC) channel encoding method for use in a wireless communications system. A communication device encodes an input bit sequence by using a LDPC matrix, to obtain an encoded bit sequence for transmission. The LDPC matrix is obtained based on a lifting factor Z and a base matrix. Embodiments of the application provide eight particular designs of the base matrix. The encoding method provided in the embodiments of the application can be used in various communications systems including the fifth generation (5G) telecommunication systems, and can support various encoding requirements for information bit sequences with different code lengths.

Abstract: Certain aspects of the present disclosure provide an efficiently decodable QC-LDPC code which is based on a base matrix, the base matrix being formed by columns and rows, the columns being dividable into one or more columns corresponding to punctured variable nodes and columns corresponding to non-punctured variable nodes. Apparatus at a transmitting side includes a encoder configured to encode a sequence of information bits based on the base matrix. Apparatus at a receiving side configured to receive a codeword in accordance with a radio technology across a wireless channel. The apparatus at the receiving side includes a decoder configured to decode the codeword based on the base matrix.

Abstract: A method for generating a code, a method for encoding and decoding data, and an encoder and a decoder performing the encoding and decoding are disclosed. In an embodiment, a method for lifting a child code from a base code for encoding and decoding data includes determining a single combination of a circulant size, a lifting function, and a labelled base matrix PCM according to an information length and a code rate using data stored in a lifting table. The lifting table was defined at a code generation stage. The method also includes calculating a plurality of shifts for the child code. Each shift is calculated by applying the lifting function to the labelled base matrix PCM with a defined index using the circulant size and using the derived child PCM to encode or decode data.

Abstract: A base matrix of a rate-adaptive irregular QC-LDPC code is provided, the base matrix being formed by columns and rows having entries representing circulant submatrices. The columns of the base matrix are divided into at least one or more higher weight first columns and lower weight second columns and the rows of the base matrix are divided into first high weight rows corresponding to the high rate mother code and second low weight rows corresponding to the extension part related to the lower rate codes. A first submatrix formed by an intersection of entries of the second columns and entries of the first and the second rows is divided into first quadratic submatrices, wherein at most one entry in each column of each first submatrix and/or at most one entry in each row of each first submatrix is labelled.

Abstract: This application discloses an encoding method, an apparatus, a communications device, and a communications system. The method includes: encoding an input bit sequence by using a low density parity check LDPC matrix, where the LDPC matrix is obtained based on a lifting factor Z and a base matrix, and the base matrix includes a row 0 to a row 4 and a column 0 to a column 26 in one of matrices shown in FIG. 3b-1A to FIG. 3b-8B, or the base matrix includes a row 0 to a row 4 and some of a column 0 to a column 26 in one of matrices shown in FIG. 3b-1A to FIG. 3b-8B. According to the encoding method, the apparatus, the communications device, and the communications system, channel coding requirements can be met.

Abstract: Embodiments of this application disclose provides a low density parity check (LDPC) channel encoding method for use in a wireless communications system. A communication device encodes an input bit sequence by using a LDPC matrix, to obtain an encoded bit sequence for transmission. The LDPC matrix is obtained based on a lifting factor Z and a base matrix. Embodiments of the application provide eight particular designs of the base matrix. The encoding method provided in the embodiments of the application can be used in various communications systems including the fifth generation (5G) telecommunication systems, and can support various encoding requirements for information bit sequences with different code lengths.

Abstract: A method for generating a code, a method for encoding and decoding data, and an encoder and a decoder performing the encoding and decoding are disclosed. In an embodiment, a method for lifting a child code from a base code for encoding and decoding data includes determining a single combination of a circulant size, a lifting function, and a labelled base matrix PCM according to an information length and a code rate using data stored in a lifting table. The lifting table was defined at a code generation stage. The method also includes calculating a plurality of shifts for the child code. Each shift is calculated by applying the lifting function to the labelled base matrix PCM with a defined index using the circulant size and using the derived child PCM to encode or decode data.

Abstract: A decoding method, an encoding method, a decoder and an encoder are disclosed. In an embodiment the decoding method includes receiving, at a receiver of a receiving side, signals from a transmitting side, the signals including a code word and decoding, at a decoder of the receiving side, the code word using a low density parity check (LDPC) code in which each n adjacent rows, n>1, in an extension part of a base parity check matrix (PCM) are orthogonal except for punctured information columns.

Abstract: A method for generating a code, a method for encoding and decoding data, and an encoder and a decoder performing the encoding and decoding are disclosed. In an embodiment, a method for lifting a child code from a base code for encoding and decoding data includes determining a single combination of a circulant size, a lifting function, and a labelled base matrix PCM according to an information length and a code rate using data stored in a lifting table. The lifting table was defined at a code generation stage. The method also includes calculating a plurality of shifts for the child code. Each shift is calculated by applying the lifting function to the labelled base matrix PCM with a defined index using the circulant size and using the derived child PCM to encode or decode data.