Abstract: Various aspects of the present disclosure generally relate to wireless communication. In some aspects, a network node may transmit, to multiple user equipments (UEs), respective over-the-air digital pre-distortion (OTA-DPD) training signals, and may receive, from the multiple UEs, multiple PA model reports. The network node may perform a DPD correction procedure in accordance with the multiple PA model reports. Additionally or alternatively, the network node may receive, from a UE, a first and second PA model report in accordance with a first and second OTA-DPD training signal, respectively. The first PA model report may indicate a set of data bits and a first set of parity bits and the second PA model report may indicate a second set of parity bits that enable information associated with the second PA model report to be decoded using at least the second set of parity bits and the set of data bits.

Abstract: Various aspects of the present disclosure generally relate to wireless communication. In some aspects, a user equipment (UE) may receive, from a network node, configuration information indicating a first grouping of a first subset of a plurality of transmission streams with a first subset of a plurality of antennas at the UE and a second grouping of a second subset of the plurality of transmission streams with a second subset of the plurality of antennas. The UE may receive a communication via the plurality of transmission streams, demodulate the first subset of the plurality of transmission streams received by the first subset of the antennas based on the configuration information, and demodulate the second subset of the plurality of transmission streams received by the second subset of the antennas based on the configuration information. Numerous other aspects are described.

Abstract: Methods, systems, and devices for wireless communication are described. A user equipment may transmit, to a serving network entity, a message requesting a quantity of transmission layers scheduled to be used by one or more neighbor network entities for communication within a communication resource. In response, the user equipment may receive, from the serving network entity, an indication of the quantity of transmission layers scheduled to be used by one or more neighbor network entities for communication within the communication resource. The user equipment may apply the quantity of transmission layers for the communication resource to a rank-aware interference rejection algorithm and may use the rank-aware interference rejection algorithm to improve a noise covariance matrix estimation. When receiving a downlink message, from the serving network entity via the communication resource, the improved noise covariance matrix estimation may be used to mitigate interference at the communication resource.

Abstract: This disclosure provides methods, components, devices and systems for signal layer combination precoding indication signaling. For example, a network entity may precode a set of layers of a signal on a per precoding group basis in accordance with one or more layer combination configurations. Each precoding group of a set of precoding groups associated with the per precoding group basis may include a respective subset of the plurality of layers. The network entity may then transmit, to a user equipment (UE), the precoded set of layers of the signal, that were precoded by the network entity, and an indication of the one or more layer combination configurations used to precode the plurality of layers on the per precoding group basis.

Abstract: Methods, systems, and devices for wireless communications are described. A network entity may transmit, to a set of multiple user equipments (UEs), an indication of a wireless communication resource to be used by the set of multiple UEs for concurrent transmission of compressed gradient data associated with a machine learning (ML) model. The network entity may receive, via the wireless communication resource, a first signal including an over-the-air (OTA) summation of the compressed gradient data. The compressed gradient data may be compressed by multiplication of gradient data by a linear compression matrix associated with a training procedure for the ML model. The network entity may transmit, to the set of multiple UEs, a second signal including the OTA summation of the compressed gradient data or indicating an update to the ML model based on the compressed gradient data.

Abstract: Certain aspects of the present disclosure provide techniques for channel equalization. An example method includes sending a first indication of a capability of the UE to use decentralized Tomlinson-Harashima Precoding (dTHP), receiving a configuration to use at least one of: centralized Tomlinson-Harashima Precoding (cTHP) to receive one or more first transmissions via one or more first resources; the dTHP to receive the one or more first transmissions via the one or more first resources; or the dTHP to receive one or more second transmissions via one or more second resources that are different than the one or more first resources.

Abstract: Various aspects of the present disclosure generally relate to wireless communication. In some aspects, a user equipment (UE) may receive, from a network node, a request for a plurality of frequency-dependent residual side band (FDRSB) reports. The UE may transmit, to the network node, a first report including a full FDRSB model associated with a first FDRSB estimation. The UE may transmit, to the network node, a second report including a partial FDRSB model associated with a second FDRSB estimation. Numerous other aspects are described.

Abstract: Various aspects of the present disclosure generally relate to wireless communication. In some aspects, a user equipment (UE) may receive configuration information indicating a set of control resource sets (CORESETs), wherein the configuration information indicates sets of allocation properties associated with respective CORESETs included in the set of CORESETs. The UE may receive a physical downlink control channel (PDCCH) communication based at least in part on monitoring resources associated with a CORESET included in the set of CORESETs. Numerous other aspects are described.

Abstract: Certain aspects of the present disclosure provide techniques for wireless communications. An example method includes receiving, from a network entity, transmit antenna correlation information associated with the network entity; performing an equalization operation based at least in part on one or more decimation factors, the one or more decimation factors being associated with the transmit antenna correlation information; and communicating with the network entity in accordance with the equalization operation.

Abstract: Methods, systems, and devices for wireless communications are described. A component of a user equipment (UE) may experience non-linearity behavior. Using a digital post distortion (DPOD) process, the UE may estimate non-linear distortion caused by the non-linearity behavior and may use the estimated non-linear distortion to cancel or mitigate distortion at a received downlink signal. To improve the non-linear distortion estimation, a network entity may determine and transmit to the UE recommended kernel set sizes to use at a DPOD algorithm of the DPOD process. The recommended kernel set sizes may be based on a quantity of receive antenna in use at the UE and an instantaneous SNR measurement at the UE. The UE may utilize one of the recommended kernel set sizes in performing the DPOD process on a received downlink signal to mitigate or cancel the non-linear distortion.

Abstract: Methods, systems, and devices for wireless communications are described. A network entity may obtain from a user equipment (UE), a report that indicates one or more signal quality values associated with a set of communication chains of the UE. The network entity may perform a frequency domain residual sideband (FDRSB) identification procedure to determine an FDRSB correction table associated with a set of frequency bands and the set of communication chains. The network entity may output to the UE a control message indicating the FDRSB correction table that identifies which one or more frequency bands of the set of frequency bands and which one or more communication chains of the set of communication chains are associated with FDRSB correction. The UE may demodulate one or more messages via at least one communication chain of the set of communication chains based on FDRSB correction performed using the FDRSB correction table.

Abstract: Certain aspects of the present disclosure provide techniques for wireless communications. An example method includes identifying a first value for a scaling factor associated with scaling a gradient indication, wherein the gradient indication is associated with communicating gradient information for federated learning at the UE; transmitting, to a network entity, the first value; receiving, from the network entity, a second value for the scaling factor; and transmitting, to the network entity, the gradient indication using the second value for the scaling factor.

Abstract: Various aspects of the present disclosure generally relate to wireless communication. In some aspects, a network node may transmit, to multiple user equipments (UEs), respective over-the-air digital pre-distortion (OTA-DPD) training signals, and may receive, from the multiple UEs, multiple PA model reports. The network node may perform a DPD correction procedure in accordance with the multiple PA model reports. Additionally or alternatively, the network node may receive, from a UE, a first and second PA model report in accordance with a first and second OTA-DPD training signal, respectively. The first PA model report may indicate a set of data bits and a first set of parity bits and the second PA model report may indicate a second set of parity bits that enable information associated with the second PA model report to be decoded using at least the second set of parity bits and the set of data bits.

Abstract: Certain aspects of the present disclosure provide techniques for wireless communications. An example method includes receiving, from a set of nodes, a set of sign indications indicating a set of signs for a set of values of a gradient associated with federated learning at the set of nodes; transmitting, to the set of nodes, a first decision regarding a selected value of the gradient in accordance with the set of sign indications; receiving, from the set of nodes, a set of second indications of a direction of the respective value of each node of the set of nodes indicating whether the respective value is greater than the selected value as indicated by the first decision or lesser than the selected value as indicated by the first decision; and transmitting a second decision indicating an update to the selected value in accordance with the set of second indications.

Abstract: Certain aspects of the present disclosure provide techniques for wireless communications. An example method includes transmitting, to a set of nodes, first information associated with a first model parameter and a first gradient step value, the first information being associated with federated learning at the set of nodes; receiving, from the set of nodes, one or more first gradient indications for the first model parameter; and transmitting, to the set of nodes, second information associated with the first model parameter and a second gradient step value different from the first gradient step value, wherein: the second gradient step value is based at least in part on a first voting reliability associated with the first model parameter, and the first voting reliability is based at least in part on the one or more first gradient indications from the set of nodes and the first set of receive antennas.

Abstract: Various aspects of the present disclosure generally relate to wireless communication. Some aspects more specifically relate to dynamically selecting frequency resources to convey a phase tracking reference signal (PTRS). In some aspects, a network node may identify that a user equipment (UE) supports a PTRS that is contiguous over time resources and frequency resources. The network node may inform the UE of possible frequency resources for PTRS allocation. The UE may identify a frequency resource, from among the possible frequency resources, having a corresponding channel that can best support PTRS transmission. The UE may report the frequency resource to the network node, and the network node may transmit the PTRS, such as a contiguous PTRS, in accordance with the reported frequency resource.

Abstract: A device for federated machine learning, comprising: a communication system; one or more memories configured to store: values of model parameters of a machine learning (ML) model, and phase values; and one or more processors are configured to: apply the ML model, using the values of the model parameters, to model input data to determine model output data; determine an error function based on the model output data and expected output data; calculate a gradient of the error function; generate a plurality of frequency-domain values based on the gradient; generate modified frequency-domain values based on the phase values and the frequency-domain values; generate a time-domain digital signal based on the modified frequency-domain values; wherein the communication system is configured to transmit an analog radio frequency (RF) signal based on the time-domain digital signal.

Abstract: Methods, systems, and devices for wireless communications are described. A network entity may indicate to a user equipment (UE) (e.g., in a downlink control message) a capability of the network entity to disable power amplifiers for downlink transmissions. In response to the capability indication, the UE may estimate and report to the network entity the delay spread and antenna array loss threshold of a downlink signal. Based on the reported delay spread and antenna array loss threshold, the network entity may determine a quantity of power amplifiers that may be disabled for downlink transmission without reducing the received signal power at the UE below a minimum threshold. The power amplifiers that the network entity disables may be the power amplifiers with input powers less than a threshold (e.g., less than a highest input power by a threshold).

Abstract: Various aspects of the present disclosure generally relate to wireless communication. Particular aspects of the subject matter described in this disclosure can be implemented to realize one or more of the following potential advantages. In some examples, the described techniques can be used to improve latency and reduce power consumption of a DPoD process performed by the network node. For example, in association with identifying the one or more kernels for the subset of candidate kernels, the network node may consume a reduced amount of power and may identify the one or more kernels with reduced latency, when compared with testing each candidate kernel. Additionally or alternatively, the UE may conserve power resources by using PAs of the UE with reduced power (with an increased NL region) in association with the improved latency of the DPoD process performed by the network node.

Abstract: Methods, systems, and devices for wireless communications are described. A network entity may indicate to a user equipment (UE) (e.g., in a downlink control message) a capability of the network entity to disable power amplifiers for downlink transmissions. In response to the capability indication, the UE may estimate and report to the network entity the delay spread and antenna array loss threshold of a downlink signal. Based on the reported delay spread and antenna array loss threshold, the network entity may determine a quantity of power amplifiers that may be disabled for downlink transmission without reducing the received signal power at the UE below a minimum threshold. The power amplifiers that the network entity disables may be the power amplifiers with input powers less than a threshold (e.g., less than a highest input power by a threshold).