Huawei Technologies Co., Limited

Wuxian Shi

Yiqun Ge

Wen Tong

Qifan Zhang

Tao Huang

Ran Zhang

Nan Cheng

Lan Hu

Louis-Philippe Hamelin

Ingmar Land

Gaoning He

Jean-Claude Belfiore

A wireless network can generate candidate signal configurations for physical transmissions to or from a user equipment (UE) in a radio environment. The generation of candidate signal configurations can be performed using a first neural network that is associated with the UE. These signal configurations can then be evaluated using a second neural network that is associated with the radio environment. The second neural network can be trained using measurements from previous physical transmissions in the radio environment. The trained second neural network generates a reward value that is associated with the candidate signal configurations. The first neural network is then trained using the reward values from the second neural network to produce improved candidate signal configurations. When a signal configuration that produces a suitable reward value is generated, this signal configuration can be used for the physical transmission in the radio environment.

Systems and methods for wireless signal configuration by a neural network

Methods and apparatuses for feature-driven communications are described. A set of features describing an observed subject is transmitted by a transmitting electronic device (ED) to a base station (BS). The BS translates the received features to another set of transmission features to be transmitted to a receiving ED. The receiving ED recovers information about the subject from the features received from the BS.

Methods and systems for relaying feature-driven communications

Methods and systems are described which use a sensing system in cooperation with a wireless communication system. Coordinate information (which may be from the sensing system, from an electronic device, or from a network-side device) and signal-related information (which may be from the wireless system) are associated with each other. The associated information may be used for wireless communication management, such as beam management operations, among others.

Methods and apparatuses using sensing system in cooperation with wireless communication system

A transmitter and receiver are provided for communication over a noisy channel in a wireless communications system. The transmitter and receiver use polar coding to provide reliability of data transmission over the noisy wireless channel. In addition, signature bits are inserted in some unreliable bit positions of the polar code. For a given codeword, the receiver with knowledge of the signature can more effectively decode the codeword. Cyclic redundancy check (CRC) bits may also included in the input vector to assist in decoding.

Signature-enabled polar encoder and decoder

A method and apparatus for encoding source information for transmission over a transmission channel is disclosed. The method involves causing a source encoder to generate a plurality of feature probability distributions representing aspects of the source information. The method also involves receiving the plurality of feature probability distributions at an input of a distribution channel encoder, the distribution channel encoder being implemented using a polarization stream network. The method further involves causing the distribution channel encoder to transform the plurality of feature probability distributions into a dimension-extended output plurality of distribution codewords for transmission over the transmission channel. Methods and apparatus for decoding the output plurality of distribution codewords to regenerate the source information is also disclosed.

Encoding and decoding source information using a distribution channel encoder and decoder

A multi-beam transmission method is provided for transmitting using an N-beam transmitter to a receiver having K receive beams. In the transmitter, a non-linear encoder implemented by a machine learning block, and a linear encoder are trained using gradient descent back propagation that relies on feedback from the receiver. For each input to be transmitted the machine learning block is used to process the input to produce N/K sets of L outputs. The linear encoder is used to perform linear encoding on each set of L outputs to produce a respective set of K outputs so as to produce N/K sets of K encoded outputs overall and N encoded outputs overall. One of the N/K sets of K outputs from each set of K beams. In order to allow a generalization to different SNRs, and to allow for time varying channel conditions, the non-linear layers, implemented in the non-linear encoder, are responsible for extracting features and regenerating the features, while the linear layers implemented in the linear encoder, are responsible for more generalization.

Hybrid message-passing-algorithm-based deep learning neural network for a massive beam-forming system

A method and apparatus for multiple-access wireless transmission is disclosed. The method involves mapping a plurality of signals onto a multi-dimensional non-Gaussian source manifold, the plurality of signals including signals targeted for transmission to a plurality of receivers. The method also involves transforming the source manifold into a multi-dimensional target manifold using a polarization stream network. The method further involves generating a multiple-access transmission waveform for transmission to the plurality of receivers, the multiple-access transmission waveform being based on the target manifold.

Multiple access wireless communications using a non-gaussian manifold

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.

Method to generate ordered sequence for polar codes

A polarization stream architecture is described. A transmitter may implement a reverse polarization stream to shape a first source signal in a first signal space to a first target signal in a second signal space. The reverse polarization stream is implemented as a cascade of reverse polarization steps. Each reverse polarization step includes a shuffle function, a split function, a scaling function and an offset function. Machine-learning techniques may be used to implement the scaling function and the offset function. A receiver may implement a polarization stream to recover the source signal.

Method and apparatus for wireless communication using polarization-based signal space mapping

Methods and apparatuses for measuring a distance between two signal distributions in a common signal space are described. A measurement network is configured to receive first and second signal distributions as input and output a measurement of a Wasserstein distance between the first and second signal distributions. The measurement network may be implemented using a neural network.

Neural-network-based distance metric for use in a communication system

A first error-detecting code (EDC) is computed based on a first segment of a block of information that is to be encoded, and a second EDC is computed based on at least a second segment of the block of information. The first EDC is masked with a first masking segment and the second EDC with a second masking segment to generate a first masked EDC and a second masked EDC. The first masking segment and the second masking segment are associated with a target receiver of the block of information. A codeword is generated based on a code and an input vector that includes the first segment, the first masked EDC, the second segment, and the second masked EDC. This type of coding could be useful to support early termination of blind detection at a decoder, for example.

Apparatus and methods for error detection coding

Methods and apparatuses for feature-driven machine-to-machine communications are described. At a feature encoder, features are extracted from sensed raw information, to generate features that compress the raw information by a compression ratio. The feature encoder implements a probabilistic encoder to generate the features, each feature providing information about a respective probability distribution that each represents one or more aspects of the subject. The probabilistic encoder is designed to provide a compression ratio that satisfies a predetermined physical channel capacity limit for a transmission channel. The features are transmitted over the transmission channel.

Methods and apparatuses for feature-driven machine-to-machine communications

Coding sub-channel selection involves, in an embodiment, determining, from sub-channels that are defined by a code and that have associated reliabilities for input bits at input bit positions, a first number of the sub-channels to carry bits that are to be encoded. A second number of the sub-channels, greater than the first number, are selected. The second number of sub-channels are selected to provide exactly the first number sub-channels to be available to carry the bits that are to be encoded.

Growing community of inventors

Kanata, Canada

Wuxian Shi