Abstract: Systems and techniques for attention calculation optimization are described. An attention calculation system identifies tensor dimensions based on a characteristic of a tensor multiplication engine. In some examples, the tensor dimensions are matrix dimensions, for instance if the characteristic indicates that the tensor multiplication engine is optimized for matrix multiplication. The attention calculation system groups at least a subset of query data into at least one query tensor having the tensor dimensions. The attention calculation system groups at least a subset of key data into at least one key tensor having the tensor dimensions. The attention calculation system determines, using the tensor multiplication engine, a tensor multiplication including the at least one query tensor and the at least one key tensor to generate output data.

Abstract: Techniques are described for wireless communication. In one method, a user equipment (UE) receives a timing synchronization signal (TSS) and a physical broadcast channel (PBCH), with the TSS based at least in part on a timing of the TSS within a broadcast channel transmission time interval (BCH TTI); determines the timing of the TSS within the BCH TTI; and demodulates the PBCH based at least in part on the TSS. In another method, a base station allocates resources for a TSS and a PBCH within a BCH TTI; determines the TSS based at least in part on a timing of the TSS within the BCH TTI; and transmits, on the resources allocated for the TSS and the PBCH, the TSS and the PBCH, with the TSS transmitted as a demodulation reference signal (DMRS) for the PBCH on at least one port used to transmit the TSS and the PBCH.

Abstract: The present disclosure relates to wireless communication in mmW networks. The apparatus may be a base station. The apparatus may be configured to determine a first symbol index and a second symbol index associated with downlink resources allocated to a UE. The first symbol index may indicate when the downlink resources begin in a subframe, and the second symbol index may indicate when the downlink resources end in the subframe. The apparatus may be configured to transmit an indication of the first symbol index and the second symbol index to the UE.

Abstract: Methods, systems, and devices for wireless communication are described. In some wireless systems, a user equipment (UE) may access a network or cell using a random access (RACH) procedure. A base station may allocate dedicated RACH resources and common RACH resources for the UE to transmit a RACH preamble message on and may convey a configuration for the UE. In some cases, the UE may perform multiple transmissions of the RACH preamble message in the dedicated RACH resources, and based on the configuration may determine whether to perform a single or multiple transmissions of the RACH preamble message in the common RACH resources. The UE may transmit the messages using the same or different uplink transmit beams, and the base station may receive the messages using the same or different uplink reception beams. These multiple transmissions may reduce latency and improve reliability of the RACH procedure.

Abstract: Methods, systems, and devices for wireless communication are described. The methods, systems, and devices provide for identifying tone spacing for transmission or reception of signals. The identified tone spacing may vary depending on the transmission or reception spectrum band or signal type. Using the identified tone spacing, a number of repetitions or a number of symbols for transmission or receiver algorithm of a signal may be determined.

Abstract: Techniques are described for wireless communication. In one method, a user equipment (UE) receives a timing synchronization signal (TSS) and a physical broadcast channel (PBCH), with the TSS based at least in part on a timing of the TSS within a broadcast channel transmission time interval (BCH TTI); determines the timing of the TSS within the BCH TTI; and demodulates the PBCH based at least in part on the TSS. In another method, a base station allocates resources for a TSS and a PBCH within a BCH TTI; determines the TSS based at least in part on a timing of the TSS within the BCH TTI; and transmits, on the resources allocated for the TSS and the PBCH, the TSS and the PBCH, with the TSS transmitted as a demodulation reference signal (DMRS) for the PBCH on at least one port used to transmit the TSS and the PBCH.

Abstract: A device for processing image data is disclosed. The device can obtain a radar point cloud and one or more frames of camera data. The device can determine depth estimates of one or more pixels of the one or more frames of camera data. The device can generate a pseudo lidar point cloud using the depth estimates of the one or more pixels of the one or more frames of camera data, wherein the pseudo lidar point cloud comprises a three-dimensional representation of at least one frame of the one or more frames of camera data. The device can determine one or more object bounding boxes based on the radar point cloud and the pseudo lidar point cloud.

Abstract: Methods, systems, and devices for wireless communication are described. A network device, such as a base station, may transmit a request message to a user equipment (UE). The request message may include a request for the UE to transmit a set of sounding reference signals (SRSs). The set of SRSs may include two (or more) beamformed signals. The network device may receive the set of SRSs according to the request message. The network device may identify, based on a co-phasing parameter associated with the two (or more) beamformed signals, an antenna port precoder configuration to use for communicating with the UE.

Abstract: In some aspects, a system may receive, from a first one-dimensional radar array, first information based at least in part on first reflections associated with an azimuthal plane. The system may further receive, from a second one-dimensional radar array, second information based at least in part on second reflections associated with an elevation plane. Accordingly, the system may detect an object based at least in part on the first information and may determine an elevation associated with the object based at least in part on the second information. Numerous other aspects are described.

Abstract: According to some aspects of the disclosure, techniques for compression techniques for the radar data that can be used in real-time applications for automated or self-driving vehicles. One or more compression techniques can be selected and/or configured based on information regarding operational conditions provided by a central (vehicle) computer. Operational conditions can include environmental data (e.g., weather, traffic), processing capabilities, mode of operation, and more. Compression techniques can facilitate transport of compressed radar data from a radar sensor to the central computer for processing of the radar data for object detection, identification, positioning, etc.

Abstract: In some aspects, a device may receive point data associated with a cell of an occupancy grid for controlling a vehicle. The device may determine, based on the point data, a characteristic of the cell that is associated with an occupancy probability of the cell, wherein the occupancy probability is determined according to a first technique based on the point data. The device may configure, based on the characteristic, the occupancy probability for the cell, within the occupancy grid, according to a second technique. Numerous other aspects are described.

Abstract: Backhaul resources may be allocated to different wireless communication links between different base stations, and local redistribution of resources among base stations may be utilized to account for variations in signal quality and/or variations in traffic experienced by different nodes of the backhaul network. A first access node function (ANF) may determine a need for additional backhaul resources, and may transmit a request message to one or more user equipment functions (UEFs) for additional backhaul resources. A UEF that receives the request message may forward the request to an associated second ANF. The second ANF, if it has backhaul resources available that the first ANF can use, may send a response to the first ANF, via the associated UEF, and the first ANF may use the additional resources that were originally allocated to the second ANF.

Abstract: Methods, systems, and devices for wireless communication are described. Signaling of a radio frequency (RF) band associated with a synchronization signal (SS) block detected by a user equipment (UE) may be transmitted to a base station. For example, a base station may transmit a set of SS blocks to multiple UEs, where the SS blocks are frequency division multiplexed such that each SS block is transmitted in a respective RF band. A UE may receive one or more of the transmitted SS blocks and determine a preferred SS block from the received SS blocks. The UE may in turn transmit an message to the base station that indicates the RF band of the preferred SS block. In some cases, the base station may optionally transmit a request to the UE to transmit the indication of the RF band.

Abstract: Disclosed are techniques for fusing camera and radar frames to perform object detection in one or more spatial domains. In an aspect, an on-board computer of a host vehicle receives, from a camera sensor of the host vehicle, a plurality of camera frames, receives, from a radar sensor of the host vehicle, a plurality of radar frames, performs a camera feature extraction process on a first camera frame of the plurality of camera frames to generate a first camera feature map, performs a radar feature extraction process on a first radar frame of the plurality of radar frames to generate a first radar feature map, converts the first camera feature map and/or the first radar feature map to a common spatial domain, and concatenates the first radar feature map and the first camera feature map to generate a first concatenated feature map in the common spatial domain.

Abstract: Backhaul resources may be allocated to different wireless communication links between different base stations, and local redistribution of resources among base stations may be utilized to account for variations in signal quality and/or variations in traffic experienced by different nodes of the backhaul network. A first access node function (ANF) may determine a need for additional backhaul resources, and may transmit a request message to one or more user equipment functions (UEFs) for additional backhaul resources. A UEF that receives the request message may forward the request to an associated second ANF. The second ANF, if it has backhaul resources available that the first ANF can use, may send a response to the first ANF, via the associated UEF, and the first ANF may use the additional resources that were originally allocated to the second ANF.

Abstract: Techniques are described for wireless communication. In one method, a user equipment (UE) receives a timing synchronization signal (TSS) and a physical broadcast channel (PBCH), with the TSS based at least in part on a timing of the TSS within a broadcast channel transmission time interval (BCH TTI); determines the timing of the TSS within the BCH TTI; and demodulates the PBCH based at least in part on the TSS. In another method, a base station allocates resources for a TSS and a PBCH within a BCH TTI; determines the TSS based at least in part on a timing of the TSS within the BCH TTI; and transmits, on the resources allocated for the TSS and the PBCH, the TSS and the PBCH, with the TSS transmitted as a demodulation reference signal (DMRS) for the PBCH on at least one port used to transmit the TSS and the PBCH.

Abstract: Methods, systems, and devices for wireless communication are described. In aspects of the present disclosure, a user equipment (UE) may report metrics (e.g., received signal power, beam identifier) about synchronization signal (SS) beams using the same (e.g., or a similar) framework that is used for channel state information reference signal (CSI-RS) reporting. Because SSs are intended to be broadcast across a wide coverage area in a beamformed manner, the SSs represent a promising complement to existing beam management techniques. Accordingly, beam management may be achieved at least in part based on reporting one or more metrics of beamformed SSs through a channel feedback framework.

Abstract: Systems, methods, and apparatuses for performing a beam recovery procedure using a second CC are disclosed. A UE may perform a beam recovery procedure using two component carriers (CCs): for example, first component carrier may be a beam formed millimeter wave (MMW) carrier having a beam recovery procedure and second component carrier may be an assisting carrier such as a sub-6 GHz carrier or a different MMW carrier. In a first example, a UE may trigger beam recovery for first component carrier (on second component carrier), generate a beam measurement report, and transmit the beam report on resources allocated for uplink transmission on second component carrier. In a second example, a new scheduling request (SR) may be defined on first component carrier for second component carrier beam recovery. In a third example, RACH resources or procedures on second component carrier may be used to perform the beam recovery for first component carrier.

Abstract: Certain aspects of the present disclosure provide techniques for beam refinement. A base station provides information regarding what transmit beams are used to transmit different reference signal (RS) resources. In some cases, the information indicates whether a same transmit beam is used to transmit a set of RS resources (e.g., symbol, sub-symbol, or frequency resources) or whether different transmit beams are used to transmit the set of RS resources.

Abstract: Disclosed are techniques for employing deep learning to analyze radar signals. In an aspect, an on-board computer of a host vehicle receives, from a radar sensor of the vehicle, a plurality of radar frames, executes a neural network on a subset of the plurality of radar frames, and detects one or more objects in the subset of the plurality of radar frames based on execution of the neural network on the subset of the plurality of radar frames. Further, techniques for transforming polar coordinates to Cartesian coordinates in a neural network are disclosed. In an aspect, a neural network receives a plurality of radar frames in polar coordinate space, a polar-to-Cartesian transformation layer of the neural network transforms the plurality of radar frames to Cartesian coordinate space, and the neural network outputs the plurality of radar frames in the Cartesian coordinate space.