Abstract: Embodiments of the present disclosure propose a solution for availability of system broadcasted information such as GNSS information which can be used for deriving time and/or frequency domain compensation. According to embodiments of this present disclosure, a validity timer for system broadcasted information is introduced in both UE and network to guarantee the system broadcasted information availability status is synchronized in two nodes, and the system broadcasted information is valid to pre-compensate in time and/or frequency domain when the timer is running. In this way, the UE can be scheduled with correct timing and frequency estimation in UL transmission.

Abstract: A semiconductor device includes a fin structure. A source/drain region is formed on the fin structure. A first gate structure is disposed over the fin structure. A source/drain contact is disposed over the source/drain region. The source/drain contact has a protruding segment that protrudes at least partially over the first gate structure. The source/drain contact electrically couples together the source/drain region and the first gate structure.

Abstract: A federated learning method includes: providing importance parameters and performance parameters by client devices respectively to a central device, performing a training procedure by the central device, wherein the training procedure includes: selecting target devices from the client devices according to a priority order associated with the importance parameters, dividing the target devices into training groups according to a similarity of the performance parameters, notifying the target devices to perform iterations according to the training groups respectively to generate trained models, transmitting the trained models to the central device, and updating a global model based on the trained models, performing the training procedure again or outputting the global model to the client devices based on a convergence value of the global model and the number of times of performing the training procedure.

Abstract: The technology of this application relates to a frontlight module and a display apparatus. The frontlight module is disposed on a side of a display panel. The frontlight module includes a light source, a light guide plate, and light guide dots. The light guide plate includes a first surface and a second surface that are disposed opposite to each other. The display panel is disposed facing the second surface. The light source is disposed on a side surface of the light guide plate. A plurality of light guide dots are disposed on the first surface or the second surface of the light guide plate. Each light guide dot has a light guide surface disposed at an angle with respect to a surface of the light guide plate. Light is fully reflected and/or refracted on the light guide surfaces to propagate to the display panel.

Abstract: A method of performing automatic tuning on a deep learning model includes: utilizing an instruction-based learned cost model to estimate a first type of operational performance metrics based on a tuned configuration of layer fusion and tensor tiling; utilizing statistical data gathered during a compilation process of the deep learning model to determine a second type of operational performance metrics based on the tuned configuration of layer fusion and tensor tiling; performing an auto-tuning process to obtain a plurality of optimal configurations based on the first type of operational performance metrics and the second type of operational performance metrics; and configure the deep learning model according to one of the plurality of optimal configurations.

Abstract: Example embodiments of the present disclosure relate to a solution for transmitting or receiving segments of information. In an aspect, a first device generates a plurality of messages to be transmitted to a second device. Each of the plurality of messages comprises a segment of information to be provided to the second device. The first device transmits a part of the plurality of messages to the second device. In response to the first device being triggered to perform an operation with a higher priority than the transmission of the plurality of messages, the first device performs the operation by interrupting the transmission of rest of the plurality of messages to be transmitted. With example embodiments of the present disclosure, it is enabled that transmission of messages conveying information segments may be interrupted to perform an operation with a higher priority.

Abstract: A shot blasting cabinet with adjustable impellers includes a cabinet, a plurality of impeller apparatuses, and a plurality of adjusting devices. A shot-blasting working line is defined between an inlet and an outlet of the cabinet. The impeller apparatuses are symmetrically disposed on upper and lower sides of the cabinet. Each impeller apparatus has a protective housing, a driving device, a pair of rotary plates, and a plurality of impeller blades. The driving device connects one of the rotary plates. The protective housing covers the pair of rotary plates and the impeller blades, and defines a shot-blasting exit surface. The shot-blasting exit surface communicates with an interior of the cabinet. The adjusting devices are used to adjust an angle and a distance of the shot-blasting exit surface relative to the shot-blasting working line.

Abstract: In some examples, for process-to-process communication, such as in function linking, a virtual channel can be provisioned to provide virtual machine to virtual machine communications. In response to a transmit request from a source virtual machine, the virtual channel can cause a data copy from a source buffer associated with the source virtual machine without decryption or encryption. The virtual channel provisions a key identifier for the copied data. The destination virtual machine can receive an indication data is available and can cause the data to be decrypted using a key accessed using the key identifier and source address of the copied data. In addition, the data can be encrypted using a second, different key for storage in a destination buffer associated with the destination virtual machine. In some examples, the key identifier and source address is managed by the virtual channel and is not visible to virtual machine or hypervisor.

Abstract: Methods and systems are provided for a spatialized display of chat messages within a messaging platform. In an example method, a computing device displays a spatial chat user interface (UI) including a virtual background and one or more spatialized avatars overlaid on the virtual background. The computing device receives chat messages sent by a participant and displays, adjacent to the spatialized avatar representing the participant, chat bubbles having a first appearance. The computing device displays chat bubbles for other participants having a second appearance. The computing device receives one or more additional chat messages sent by the participant. The computing device displays additional chat bubbles corresponding to the additional chat messages having a third appearance and updates the appearance of the initial chat bubbles to a fourth appearance.

Abstract: An apparatus having a first portion including a first front wall, a first rear wall, and a bottom wall integrally coupled to the first front wall and the first rear wall, and pivotal pin structures integrally coupled to and extending from the first rear wall. The apparatus includes a second portion having a second front wall, a second rear wall, and a top wall integrally coupled to the second front wall and the second rear wall, and pin holders integrally coupled to and extending from the second rear wall and at an offset angle with reference to the top wall. The pivotal pin structure includes a base support connected to the first rear wall and a shaft connected to the base support, and the pin holder defines an opening sized and shaped to accept the shaft. The first and second portions are sized and shaped to be pivotally movable between open and closed configurations.

Abstract: Embodiments of the present disclosure relate to devices, methods, apparatuses and computer readable storage media of signalling enabling Timing Advance (TA) exchange between User Equipment (UE) and Radio Access Network (RAN). The method comprises receiving, at a first device, a TA margin from a second device; transmitting, to the second device, information associated with a reference location of the first device at a first time point; determining a reference TA of the first device at least based on the reference location of the first device; determining an actual TA of the first device at least based on an actual location of the first device at a second time point; and determining, based on a comparison of a TA margin with a difference between the reference TA and the actual TA, whether to transmit an updated reference location of the first device to the second device. In this way, the control for uplink overhead from TA/location updates and probability of impacting downlink scheduling become flexible.

Abstract: Example embodiments of the present disclosure relate to repetition solution for transmission. In such solution, a first device determines information associated with a movement direction of a second device serving the first device relative to the first device. Then, the first device performs a transmission of the information to the second device.

Abstract: Disclosed are methods for configuring measurement gap patterns. An example method may include: receiving configuration information associated with a plurality of pre-configured measurement gap patterns; receiving index information for configuring at least one measurement gap pattern of a bandwidth part; and configuring the at least one measurement gap pattern of the bandwidth part based on at least one of the configuration information and the index information. Related apparatuses and computer readable media are also disclosed.

Abstract: An inductor device includes an 8-shaped inductor and a ring-type wire. The ring-type wire is disposed around an outer side of the 8-shaped inductor. The 8-shaped inductor includes an input terminal and a center-tapped terminal. The input terminal of the 8-shaped inductor is located on a first side of the inductor device, and the center-tapped terminal is located on a second side of the inductor device. The ring-type wire includes an input terminal and a ground terminal. The input terminal of the ring-type wire is located on the first side of the inductor device, and the ground terminal is located on the second side of the inductor device. The input terminal of the ring-type wire is coupled to the input terminal of the 8-shaped inductor.

Abstract: Embodiments of the present disclosure relate to conditional handover. A method comprises: determining, at a first device, an ordered list of target cells for a second device to initiate handover to in order based at least in part on location-related data of the second device, the second device being served in a source cell of the first device; transmitting, to the second device, first information indicating the ordered list of target cells and radio configurations associated with each of the target cells; and causing second information indicating the ordered list of target cells to be transmitted to at least one third device serving target cells in the ordered list, to request allocation of resources for the handover by the at least one third device. In this way, a chain of target cells regarding future handover is determined in a predetermined order, which avoid multiple cells prepare for a same handover, thus reducing handover preparation latency and improving the service continuity.

Abstract: A system receives a neural network model that includes asymmetric operations. Each asymmetric operation includes one or more fixed-point operands that are asymmetrically-quantized from corresponding floating-point operands. The system compiles a given asymmetric operation of the neural network model into a symmetric operation that includes a combined bias value. A compiler computes the combined bias value is a constant by merging at least zero points of input and output of the given asymmetric operation. The system then generates a symmetric neural network model including the symmetric operation for inference hardware to execute in fixed-point arithmetic.

Abstract: A liquid collecting mechanism that includes a liquid collection tube arm and a liquid collection tube head fixed at a front end of the tube arm. A sealing spacer is arranged at a joint connecting the liquid collection tube head and the liquid collection tube arm, and the liquid collection tube head is internally provided with a liquid collection cavity. An area of the liquid collection tube head located above the liquid collection cavity is a mesh collection area, and a plurality of collecting holes is provided in the mesh collection area. The mechanism includes a liquid suction tube passing through the tube arm and tube head, and the liquid suction tube has an open first end located in the liquid collection cavity and an open second end configured to connect with a pump body mechanism.

Abstract: An inductor device includes an 8-shaped inductor and a ring-type wire. The ring-type wire is disposed around an outer side of the 8-shaped inductor. The 8-shaped inductor includes an input terminal and a center-tapped terminal. The input terminal of the 8-shaped inductor is located on a first side of the inductor device, and the center-tapped terminal is located on a second side of the inductor device. The ring-type wire includes an input terminal and a ground terminal. The input terminal of the ring-type wire is located on the first side of the inductor device, and the ground terminal is located on the second side of the inductor device. The input terminal of the ring-type wire is coupled to the input terminal of the 8-shaped inductor.

Abstract: Disclosed are methods for selecting a random access procedure type. An example method may include receiving configuration information including a threshold associated with a transmission quality factor via a non-terrestrial network downlink, obtaining a value of the transmission quality factor, and selecting a 2-step random access procedure or a 4-step random access procedure based on the threshold and the obtained value of the transmission quality factor. Related apparatuses and computer readable media are also disclosed.

Abstract: A method for treating and phosphatizing a metal board without using acid includes the following steps: performing a degreasing step to remove grease and dirt from a surface of the metal board with a degreasing agent; performing a blast-peening step by blasting and peening polygon blast-peening granules on the metal board through a centrifugal impeller to remove an oxidized layer; performing a washing step to clean remaining powders from the metal board after the blast-peening step; performing a phosphatizing step to form a protective phosphate coating on the metal board; performing another washing step to wash off remaining phosphatizing agents from the metal board; performing a rustproofing step to apply a rustproofing agent on the metal board; and performing a drying step to dry the metal board.