A transmitting apparatus includes an analog transmitting unit that performs analog processing on a multiplex signal in which two or more signals are multiplexed and generates a transmission signal, and a signal multiplexing and learning unit that multiplexes the two or more signals with a neural network whose parameters have been adjusted based on the analog characteristics of the analog transmitting unit and constraints on the multiplex signal and generates the multiplex signal.

A method performed by a communication device includes generating an initial channel estimate of a channel for a current time step with a Kalman filter based on a first signal received at the communication device. The method also includes inferring, with a neural network, a residual of the initial channel estimate of the current time step. The method further includes updating the initial channel estimate of the current time step based on the residual.

A method of wireless communication by a user equipment (UE) comprises receiving, from a base station, multiple reference signals, and estimating a channel based on the received reference signals. The channel comprises multiple channel paths. The method also includes quantizing an angle of arrival (AoA) of each channel path into one of a group of quantization levels. The method further includes reporting to the base station the quantized angle of arrival, and also a delay and/or power level for the quantized angle of arrival.

Certain aspects of the present disclosure provide techniques for qualifying machine learning model-based channel state information (CSI) predictions. An example method generally includes receiving, from a network entity, a channel state information (CSI) prediction model for quantized CSI, calculating CSI based on downlink reference signal measurements, generating a quantized CSI difference value based a quantization of a difference between the calculated CSI and CSI predicted based on a CSI prediction model, and reporting, to the network entity, the calculated CSI and the quantized CSI difference value.

The embodiments of the present application provide a method and apparatus for modulation recognition of signals based on cyclic residual network, the method comprises: obtaining a signal matrix of a to-be-recognized signal, and extracting real part information and imaginary part information of the signal matrix; generating, according to extracted real part information and imaginary part information, a real-and-imaginary-part feature matrix of the to-be-recognized signal; converting, according to a preset matrix conversion method, the real-and-imaginary-part feature matrix into an amplitude-and-phase feature matrix; and inputting the amplitude-and-phase feature matrix into a pre-trained cyclic residual network to obtain a modulation mode corresponding to the to-be-recognized signal. In the embodiments of the present application, the processing of the to-be-recognized signal is simple and easy to operate, in which neither complex algorithms nor manual processing is required, the flexibility of recognition is high, and the result of modulation recognition of the to-be-recognized signal can be accurately obtained.

To obtain delay spread estimations and/or Doppler spread estimations, data representing received data is input to at least one trained model, the trained model outputting spread estimations.

A method performs eigen decomposition with an artificial deep neural network. The deep neural network receives an input covariance matrix. The deep neural network has a number of convolutional layers and also a number of pooling layers. The deep neural network predicts dominant eigen information of the input covariance matrix, after applying the convolutional layers and the pooling layers to the input covariance matrix. The input covariance matrix may be a real-valued covariance matrix or a complex-valued covariance matrix having a concatenated pair of matrices, including a first matrix of real components and a second matrix of imaginary components. The dominant eigen information may be absolute values of a pair of dominant eigen values and sign information of the pair of dominant eigen values, and/or absolute values of a pair of dominant eigen vectors and sign information of the pair of dominant eigen vectors.

To provide demapping at a receiving side, a trained model for a demapper is used to output log-likelihood ratios of received signals representing data in a multi-user transmission. Inputs for the trained model for the demapper comprise a resource grid of equalized received signals.

Various embodiments may employ neural networks at transmitting devices to compress transmit (TX) waveform distortion. In various embodiments, compressed TX waveform distortion information may be conveyed to a receiving device. In various embodiments, the signaling of TX waveform distortion information from a transmitting device to a receiving device may enable a receiving device to mitigate waveform distortion in a transmit waveform received from the transmitting device. Various embodiments include systems and methods of wireless communication by transmitting a waveform to a receiving device performed by a processor of a transmitting device. Various embodiments include systems and methods of wireless communication by receiving a waveform from a transmitting device performed by a processor of a receiving device.

Various aspects of the present disclosure generally relate to wireless communication. In some aspects, a first device may transmit a neural network capability indication that indicates a capability of the first device associated with training at least one channel state feedback (CSF) neural network for facilitating providing CSF. The first device may receive, based at least in part on the capability of the first device, a CSF neural network configuration that indicates at least one parameter associated with the at least one CSF neural network. Numerous other aspects are provided.