A method and devices for configuring a data transmission network are disclosed. The method is for configuring a data transmission network, executed by a configuration device, wherein the data transmission network comprises at least one transmitter, at least one receiver with a communication channel between the transmitter and the receiver, the method comprising: training a machine learning model of the data transmission network, wherein the machine learning model comprises at least a transmitter model including a transmitter neural network, a channel model, and a receiver model including a receiver neural network by providing a message within a sequence of messages; generating a group of transmission symbols for each message in the sequence of messages using the transmitter neural network; concatenating the groups of transmission symbols together as a sequence of transmission symbols; simulating transmission of the sequence of transmission symbols over the communication channel using the channel model to the receiver; analysing a sequence of received symbols using the reception neural network to generate a decoded message; and updating the machine learning model based on an output of said reception neural network. In this way, the machine learning model can be trained using representative sequences of message, which improves performance when deployed in a real network.

A user equipment may be configured to perform demodulator selection based on ML model coefficients trained by a base station. In some aspects, the user equipment may transmit a dynamic demodulator indication to a base station, transmit channel information to the base station, and receive, in response to the dynamic demodulator indication, updated coefficient information based on the channel information. Further, the user equipment may select a demodulator based on the updated coefficient information, and communicate with the base station via the demodulator in response to the selection of the demodulator.

A machine learning (ML) agent operates at a transmitter to optimize signals transmitted across a communications channel. A physical signal modifier modifies a physical layer signal prior to transmission as a function of a set of signal modification parameters to produce a modified physical layer signal. The ML agent parses a feedback signal from a receiver across the communications channel, and determines a present tuning status as a function of the signal modification parameters and the feedback signal. The ML agent generates subsequent signal modification parameters based on the present tuning status and a set of stored tuning statuses, thereby updating the physical signal modifier to generate a subsequent modified physical layer signal to be transmitted across the communications channel.

The present subject matter relates to a receiver including a detector for receiving a signal from a transmitter. The detector includes a set of one or more settable parameters, and circuitry configured for implementing an algorithm having trainable parameters. The algorithm is configured to receive as input information indicative of a status of a communication channel between the transmitter and the receiver and to output values of the set of settable parameters of the detector. The detector is configured to receive a signal corresponding to a message sent by the transmitter and to provide an output indicative of the message based on the received signal and the output values of the set of settable parameters of the detector.

According to the present document, a method by which a terminal receives data in a wireless communication system comprises: receiving a channel signal and a reference signal (RS) from a base station; generating a sequence by performing an operation of equalizing the RS to a channel RS; and decoding the received channel signal on the basis of the generated sequence, wherein the operation of equalizing the RS to the channel RS is based on a parameter determined according to a machine learning process.

Examples described herein include methods, devices, and systems which compensates input data for I/Q imbalance or noise related thereto to generate compensated input data. In doing such the above compensation, during an uplink transmission time interval (TTI), a switch path is activated to provide converted input data to a receiver stage including a recurrent neural network (RNN). The RNN calculates an error representative of the noise based partly on the input signal to be transmitted and a feedback signal to generate filter coefficient data associated with the I/Q imbalance. The feedback signal is provided, after processing through the receiver, to the RNN. During an uplink TTI, the converted input data is transmitted as the RF wireless transmission via an RF antenna. During a downlink TTI, the switch path is deactivated and the receiver stage receives an additional RF wireless transmission to be processed in the receiver stage.

A system, apparatus, and method are provided for performing fast re-training of fully fused neural networks configured to implement at least a portion of a transceiver. At least one of a demapping module, an equalization module, or a channel estimation module can be implemented, at least in part, using a fully fused neural network. The neural network can be trained online during operation by acquiring training data sets using a number of received frames of data. Re-training of the neural network is performed periodically to adapt the neural network to changing channel characteristics. In various embodiments, a neural demapper, a neural channel estimator, and a neural receiver are disclosed to replace or augment one or more components of the transceiver. In another embodiment, an auto-encoder can be implemented across a transmitter and receiver to replace most of the components of the transceiver, the auto-encoder being trained via an end-to-end learning algorithm.

Apparatuses, methods, and systems are disclosed for supporting JED model selection and training. One apparatus includes a processor and a transceiver that receives a configuration from a network device, said configuration indicating at least one of: a set of resources for model training, a type of intended model training, and combinations thereof. The processor selects a Joint Channel Equalization and Decoding (“JED”) model from a set of models based on the received configuration. The processor trains the selected JED model using the received configuration.

Systems and methods for classifying baseband signals include receiving, at a pre-processing stage of a neural network whose objective is modulation classification performance, a complex quadrature vector of interest including a plurality of samples of a baseband signal derived from a radio frequency signal of an unknown modulation type, providing the vector of interest to a plurality of FIR filters, each of which outputs a respective intermediate filtered version of the vector of interest, combining the outputs of two or more of the FIR filters to produce a filtered version of the vector of interest, including applying respective weightings to the outputs of the FIR filters, and providing the filtered version of the vector of interest to an analysis stage of the neural network for classification with respect to a plurality of known modulation types. The neural network may apply attention-based selection to learn the filters and respective weightings.

One or more processors control processing of radio frequency (RF) signals using a machine-learning network. The one or more processors receive as input, to a radio communications apparatus, a first representation of an RF signal, which is processed using one or more radio stages, providing a second representation of the RF signal. Observations about, and metrics of, the second representation of the RF signal are obtained. Past observations and metrics are accessed from storage. Using the observations, metrics and past observations and metrics, parameters of a machine-learning network, which implements policies to process RF signals, are adjusted by controlling the radio stages. In response to the adjustments, actions performed by one or more controllers of the radio stages are updated. A representation of a subsequent input RF signal is processed using the radio stages that are controlled based on actions including the updated one or more actions.