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.

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.

A method includes storing multiple signals received from a user equipment (UE) in a queue. The method also includes estimating a sounding reference signal (SRS) signal-to-noise-ratio (SNR) and determining a filtered SNR based on the received signals. The method also includes computing one or more features based on the filtered SNR and at least some of the received signals in the queue. The method also includes determining (i) a channel condition of the UE and (ii) a speed range of the UE based on the one or more computed features, wherein the channel condition of the UE comprises line-of-sight (LoS) or non-line-of-sight (NLoS). The method also includes determining a transmission configuration based on the channel condition of the UE and the speed range of the UE.

It is provided a method, comprising identifying a value of an onsite channel characteristic of a receive channel; requesting a neural network parameter, wherein the request comprises an indication of the onsite channel characteristic; monitoring if the neural network parameter is received in response to the request; estimating the receive channel by a neural network using the neural network parameter if the neural network parameter is received.

A pilot information symbol sending method, a channel estimation method, and a communications device. The method includes: determining, based on a discrete Fourier transform DFT matrix and a sensing matrix, a pilot information symbol corresponding to each antenna on each pilot resource; and sending a corresponding pilot information symbol on each of the pilot resources for each of the antennas; where the sensing matrix is determined through training of channel information.

In an aspect of the disclosure, a method, a computer-readable medium, and an apparatus are provided. The apparatus may receive at least two reference signals from a base station. The apparatus may determine CSI associated with at least one of the at least two reference signals. The apparatus may determine at least one parameter based on the CSI. The apparatus may transmit, to the base station, the at least one parameter and the CSI to enable a predicted CSI to be determined based on the at least one parameter and the CSI. The apparatus may receive data or control information from the base station based on predictive CSI determined by the base station using the transmitted at least one parameter and CSI.

A method for user equipment for improving a transmission efficiency on a radio channel including obtaining a predictive environmental model and predicting a channel dynamic of the radio channel based on the predictive environmental model, receiving a reference signal to measure a channel characteristic of the radio channel and calculating the channel dynamic of the radio channel based on the reference signal, and determining a deviation between the predicted channel dynamic and the calculated channel dynamic and adjusting a transmission parameter based on the deviation to improve the transmission efficiency on the radio channel.

A method may include generating an estimated time offset based on a reference signal in a communication system, and adjusting a transform window in the communication system based on the estimated time offset, wherein the estimated time offset is generated based on machine learning. Generating the estimated time offset may include applying the machine learning to one or more channel estimates. Generating the estimated time offset may include extracting one or more features from one or more channel estimates, and generating the estimated time offset based on the one or more features. Extracting the one or more features may include determining a correlation between a first channel and a second channel. The correlation may include a frequency domain correlation between the first channel and the second channel. Extracting the one or more features may include extracting a subset of a set of features of the one or more channel estimates.

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.

Systems, methods, and devices to reduce the channel estimation overhead by collecting data from many UEs and building a location-based mathematical model are disclosed. During building of the model, a reference signal is used to collect location- and signal-related data from connected UEs. Once the model is successfully built, it is then transmitted and/or downloaded to each new UE that connects to the base station. The UEs and/or the base stations then use this model to determine their own transmission parameter values. The UEs also report their location to the base stations, which use the model to estimate channel conditions and adapt transmission parameters for themselves.