Abstract: An RF transmitter with a power combiner and a differential amplifier is provided. The power combiner converts a differential output signal to a single-end output signal and transmits the single-end output signal to the antenna. The differential amplifier includes common-source input transistors, common-gate output transistors and a switch module. The common-source input transistors amplify a differential input signal and output an amplified differential signal. The common-gate output transistors, including sources electrically coupled to the common-source input transistors and drains electrically coupled to the power combiner, generate the differential output signal according to the amplified differential signal. The switch module is electrically coupled between the gates. The switch module electrically couples the gates of the common-gate output transistors if the RF transmitter is in operation and electrically isolates the gates if the RF receiver is in operation.

Abstract: An RF transmitter with a power combiner and a differential amplifier is provided. The power combiner converts a differential output signal to a single-end output signal and transmits the single-end output signal to the antenna. The differential amplifier includes common-source input transistors, common-gate output transistors and a switch module. The common-source input transistors amplify a differential input signal and output an amplified differential signal. The common-gate output transistors, including sources electrically coupled to the common-source input transistors and drains electrically coupled to the power combiner, generate the differential output signal according to the amplified differential signal. The switch module is electrically coupled between the gates. The switch module electrically couples the gates of the common-gate output transistors if the RF transmitter is in operation and electrically isolates the gates if the RF receiver is in operation.

Abstract: High-voltage semiconductor devices are provided. The high-voltage semiconductor device includes a substrate having a first conductive type and a gate region disposed on the substrate. The high-voltage semiconductor also includes a source region and a drain region disposed on two sides of the gate region respectively. The high-voltage semiconductor also includes a linear doped region disposed between the gate region and the drain region, wherein the linear doped region has a nonuniform doping depth and the first conductive type. The high-voltage semiconductor further includes a first buried layer disposed under the source region and having the first conductive type.

Abstract: A communications apparatus includes a radio transceiver and a processor. The radio transceiver receives a plurality of feedback messages from a peer device. The feedback messages are transmitted by the peer device to indicate a result of UL data transmission. The processor analyzes content of the feedback messages to determine a preferred retransmission type of the peer device and records the preferred retransmission type of the peer device in a memory device. When the radio transceiver receives a first UL grant message indicating transmission of a first UL data and the processor determines that a forthcoming first feedback message corresponding to the transmission of the first UL data is unable to be received, the processor determines whether to retransmit the first UL data according to the preferred retransmission type of the peer device.

Abstract: A method is provided for automatically renewing digital certificates in advance of their expiration in field deployed devices. The method includes generating a certificate renewal request comprising a request for at least one renewed digital certificate according to a renewal paradigm in which the at least one renewed digital certificate is generated before the at least one of the digital certificates expires, providing the certificate renewal request to the offline domain, obtaining, in the online domain from the offline domain, the at least one renewed digital certificate, and transmitting the least one renewed digital certificate to the client domain for storage in the HSM in place of the at least one of the subset of the plurality of digital certificates.

Abstract: Aspects of the disclosure provide a method for packet header compression. The method can include receiving a packet in a communication device having resources for packet compression, parsing a header of the packet to obtain header information, and determining whether to compress the header of the packet based on the header information and status of the resources for packet compression in the communication device.

Abstract: A data object update system provides a flexible framework that can be used to upgrade, renew, replace or supplement data objects that are provisioned in a large base of network-enabled devices that been deployed in the field to end users. The system has the flexibility to configure, for example, the following items, based on different requirements received from network operators: which device key and/or certificate is to be used to authenticate request messages from network-enabled devices before a specific data object update request is accepted into the system; which device identifier is to be used to authorize data object update requests; which device identifier is to be used for generating device specific data objects; and which protection mechanism is to be used to secure the delivery of data objects to network-enabled devices.

Abstract: Video description generation using neural network training based on relevance and coherence is described. In some examples, long short-term memory with visual-semantic embedding (LSTM-E) can maximize the probability of generating the next word given previous words and visual content and can create a visual-semantic embedding space for enforcing the relationship between the semantics of an entire sentence and visual content. LSTM-E can include a 2-D and/or 3-D deep convolutional neural networks for learning powerful video representation, a deep recurrent neural network for generating sentences, and a joint embedding model for exploring the relationships between visual content and sentence semantics.

Abstract: The present invention provides a FinFET device, including at least one fin structure, wherein the fin structure has a first-type well region, and a second-type well region adjacent to the first-type well region, a trench located in the fin structure and disposed between the first-type well region and the second-type well region, an insulating layer disposed in the trench, and a metal gate crossing over and disposed on the insulating layer.

Abstract: A method and apparatus is provided for managing the eligibility of data signing in an online code signing system. The method is used by a plurality of data publishers in an online code signing system. The method includes defining, by an administrator of the system, a hierarchy of a plurality of entities, and managing, by an administrator of the system, eligibility to designate at least one of a plurality of users to access the at least one configuration entity to sign the data via a plurality of accounts and eligibility to designate at least one of a plurality of managers via owner account to manage user access to sign data for at least one model entity.

Abstract: The present invention provides a FinFET device, including at least one fin structure, wherein the fin structure has a first-type well region, and a second-type well region adjacent to the first-type well region, a trench located in the fin structure and disposed between the first-type well region and the second-type well region, an insulating layer disposed in the trench, and a metal gate crossing over and disposed on the insulating layer.

Abstract: Disclosed is a manufacturing process and feature licensing system for provisioning personalized (device-unique) licenses to devices, with the following characteristics. The system is secure in that it uses a secure key wrapping mechanism to deliver the LSK to LPS. Another feature is that various network communication links are secured using standard security protocol. Further, application messages, license templates, licenses are digitally signed. The system is also flexible because it is configured to allow multiple manufacturers and to allow various feature configurations via the use of License Template. The system is also scalable, as it is possible to use multiple LPS hosts to serve multiple programming stations. The system is available in that the delegation of license signing capability from CLS to LPS eliminates the dependency on unreliable Internet connections. Redundant LPS hosts provide high level of availability required for high volume license provisioning.

Abstract: Aspects of the disclosure provide a method for reordering. The method includes receiving, by a terminal device, a stream of data units from a wireless network. The data units have respective sequence numbers. Further, the method includes reordering at a protocol layer in a protocol stack, the data units for a next process according to the sequence numbers, determining a dependency relationship of a data unit to one or more missing data units that have sequence numbers prior to the data unit, and advancing, out of order of the sequence numbers, the data unit to the next process when the data unit is independent of the one or more missing data units.

Abstract: Video description generation using neural network training based on relevance and coherence is described. In some examples, long short-term memory with visual-semantic embedding (LSTM-E) can maximize the probability of generating the next word given previous words and visual content and can create a visual-semantic embedding space for enforcing the relationship between the semantics of an entire sentence and visual content. LSTM-E can include a 2-D and/or 3-D deep convolutional neural networks for learning powerful video representation, a deep recurrent neural network for generating sentences, and a joint embedding model for exploring the relationships between visual content and sentence semantics.

Abstract: Disclosed is a manufacturing process and feature licensing system for provisioning personalized (device-unique) licenses to devices. The secure system uses a secure key wrapping mechanism to deliver the LSK to LPS. Another feature is that various network communication links are secured using standard security protocol. Application messages, license templates, licenses are digitally signed. The system is flexible, configured to allow multiple manufacturers and to allow various feature configurations via the use of License Template; scalable, as it is possible to use multiple LPS hosts to serve multiple programming stations; and available in that the delegation of license signing capability from CLS to LPS eliminates the dependency on unreliable Internet connections. Redundant LPS hosts provide high level of availability required for high volume license provisioning.

Abstract: Video highlight detection using pairwise deep ranking neural network training is described. In some examples, highlights in a video are discovered, then used for generating summarization of videos, such as first-person videos. A pairwise deep ranking model is employed to learn the relationship between previously identified highlight and non-highlight video segments. This relationship is encapsulated in a neural network. An example two stream process generates highlight scores for each segment of a user's video. The obtained highlight scores are used to summarize highlights of the user's video.

Abstract: A self-assembly coating material including carbon particles and polymer shells is provided. The polymer shells respectively cover and are bonded to the carbon particles, wherein each polymer shell has both a first functional group for adsorbing on a surface of a substrate and a second functional group for self cross-linking. The first functional groups include thiol groups. The second functional groups include epoxy groups or carboxylic groups. The self-assembly coating material can be applied to a metal substrate to form a heat dissipation layer.

Abstract: To facilitate LWA PDCP setting/scheduling, a method of providing dynamic PDCP status report polling is proposed. To enhance scheduling efficiency, the transmitter (can be eNB from DL or UE for UL) can dynamically poll PDCP status report for LWA behavior. The polling can be done by a standalone PDCP control PDU or use reserved bit in PDCP data PDU. The triggering condition for PDCP status polling includes: PDCP without poll time exceeds a first threshold, PDCP without poll bytes exceeds a second threshold, PDCP without poll PDU numbers exceeds a third threshold, a data buffer is empty, and when LWA is deactivated.

Abstract: A method, an electronic device, and a computer program product for displaying a virtual button are provided in the invention. The method includes the following steps. First, the virtual button is activated. Each interactive object displayed on a display frame of a touch screen is detected to obtain an object region occupied by each of the interactive objects, and an initial position of the virtual button on the display frame is obtained. Next, whether the initial position of the virtual button is located within the object region determined. If yes, the virtual button would be displayed at the initial position. If no, whether a non-object region satisfies a first predetermined condition is determined so as to accordingly display the virtual button at the initial position, a first designated position lies within the non-object region, or a second designated position lies within both of the object region and the non-object region.