Abstract: The present disclosure is drawn to the reduction of parameters in fully connected layers of neural networks. For a layer whose output is defined by y=Wx, where y is the output vector, x is the input vector, and W is a matrix of connection parameters, vectors uij and vij are defined and submatrices Wi,j are computed as the outer product of uij and vij, so that Wi,j=vij?uij, and W is obtained by appending submatrices Wi,j.

Abstract: A data transmission method, a terminal device, and a base station are described. The data transmission method determines, by a terminal device, a first reference signal (RS) sequence corresponding to a first time interval, where the first RS sequence is one of N RS sequences, N is a positive integer greater than 1, and the N RS sequences are generated based on at least two root sequences, sends, by the terminal device, the first RS sequence to a base station in the first time interval, determines, by the terminal device from the N RS sequences according to the first RS sequence, one RS sequence as a second RS sequence corresponding to a second time interval, where the second RS sequence is determined according to the first RS sequence, and sends the second RS sequence to the base station in the second time interval.

Abstract: A method includes obtaining, by a first device, BLUETOOTH connection information of a second device, obtaining, by the first device according to the BLUETOOTH connection information, a BLUETOOTH signal from the second device, determining, by the first device according to the obtained BLUETOOTH signal from the second device, position information of the second device, and sending, by the first device, the position information to a third device.

Abstract: Embodiments of the present disclosure provide a data transmission method, which can meet a requirement for an Ethernet network with diversified rate levels. The method includes: grouping media access control (MAC) layer data into a plurality of MAC layer data groups; allocating, according to a bandwidth required by a target MAC layer data group and a reference bandwidth of a logical channel, at least one target logical channel to the target MAC layer data group; encoding the target MAC layer data group to generate target physical layer data, where the target logical channel corresponds to the target MAC layer data group and the target physical layer data; and sending the target physical layer data and first indication information, where the first indication information is used to indicate a relationship between the target physical layer data and the target logical channel.

Abstract: A method and a terminal device for keeping a subscriber identity module card in a standby state, where the method for keeping a subscriber identity module card in a standby state includes receiving, by a terminal device, an enabling request, where the enabling request is used for requesting simultaneous standby of a first subscriber identity module card and a second subscriber identity module card; invoking a third-party invocation interface provided by an operating system of the terminal device; performing interaction with a subscriber identity module card control module, so as to complete a network access process of the first subscriber identity module card on the basis of a data file of the first subscriber identity module card; and keeping the first subscriber identity module card and the second subscriber identity module card in a standby state at the same time according to the enabling request.

Abstract: A monitoring and calibration apparatus for an optical networking device such as ROADM is provided. Reflectors are integrated into the device, for example at the ends of optical interconnect cables. The reflectors reflect light in specific monitoring wavelengths and pass other wavelengths such as those used for communication. A light source emits monitoring light which is reflected by the reflector and measured by a detector to measure the integrity of optical paths. The optical paths can include optical cables and cable connectors. Path integrity between different modules of the device can therefore be monitored. Multiple reflectors, reflecting light in different wavelengths, can be placed in series along the same optical path and used to monitor multiple segments of the path. A wavelength selective switch (WSS) of the device can be used to route monitoring light to different optical paths. The WSS also operates to route communication signals in the device.

Abstract: A method of transmitting includes categorizing a transmission between the first device and a second device as one of a plurality of transmission types, and selecting an air interface from a plurality of air interface candidates in accordance with the transmission as categorized. The method also includes sending the transmission to the second device using the selected air interface.

Abstract: The present disclosure provides example methods for changing graphics processing resolution according to a scenario. In one example, a first display scenario is determined as a scenario in which energy can be saved. Based on the determination, a graphics processing resolution of a graphics processor can be reduced. At least one target graphics frame in the first display scenario can be rendered by the graphics processor according to the reduced graphics processing resolution to obtain at least one target image frame. The at least one target image frame can be adapted according to screen display resolution, and the at least one adapted target image frame can be displayed. The present disclosure further provides a portable electronic device and a system for changing graphics processing resolution according to a scenario.

Abstract: In various aspects, an optical connector with a built-in photodetector for detecting light in the optical connector; an adapter for detachably receiving an optical connector, comprising an electrode configured to electrically contact an electrode of the optical connector; a connection apparatus comprising an arrangement of adapters for detachably receiving a plurality of optical connectors, each adapter comprising an electrode configured to electrically contact an electrode of the respective optical connector; a system with an interface configured to make an optical connection with an optical connector and a processing unit configured to receive from the connector an indication and to generate a feedback signal indicative of said indication; and/or an optical connector responsive to light to generate an electrical signal is provided.

Abstract: A method for a terminal to transmit an uplink signal includes: receiving an instruction sent by a serving base station and used for determining transmit power; selecting an uplink power control formula from a plurality of uplink power control formulas according to the instruction; obtaining transmit power of an uplink signal by using the selected uplink power control formula; and transmitting an uplink signal by using the transmit power. An uplink power control formula is selected from a plurality of uplink power control formulas according to an instruction of a serving base station, so that a terminal is not only capable of selecting, according to a receiving manner, a corresponding uplink power control formula to determine a transmit signal, but also capable of using a corresponding uplink power control formula for a different configuration parameter of a non-periodical SRS to determine suitable transmit power to transmit an uplink signal.

Abstract: An assembly with optical gain assisted optical transposer is provided. The optical transposer which optically couples a fibre array unit and a photonic integrated circuit. The optical transposer includes one or more optical gain elements which are configured to provide optical compensation, for example optical gain to mitigate optical losses associated with multistage photonic integrated devices. According to some embodiments, the optical gain element is a semiconductor optical amplifier (SOA). According to some embodiments the photonic integrated circuit is a SiPh PIC.

Abstract: Methods and systems for providing an endpoint device with access to a remote resource are disclosed. A first secure tunnel with the endpoint device is established from an intermediate device, the first tunnel terminating within a trusted execution environment (TEE) in the intermediate device. At least one credential is received within the TEE and via the first secure tunnel from the endpoint device. The at least one credential is transmitted from the intermediate device to the remote resource via a second secure tunnel, the second tunnel located between the remote resource and the intermediate device and originating within the TEE. In response to the at least one credential being accepted by the remote resource, communications between the endpoint device and the remote resource via the TEE in the intermediate device through the first and second secure tunnels are enabled.

Abstract: Embodiments are provided for uplink measurement based mechanism and control using user equipment (UE) centric sounding signals. The mechanism provides an alternative to DL-measurement dominated system control. Based on UL-measurements at TPs, the network obtains knowledge of users' channel and timing information, traffic, and interference, and is thus able to perform better control, including TP and UE clustering and optimization, and power control and link adaptation. In an embodiment method, a TP receives one-to-one mapping information indicating a plurality of UE IDs and a plurality of sounding channels assigned to the corresponding UE IDs. When the TP detects a sounding reference signal (SRS) from a UE, the TP is able to identify the UE using the detected SRS and the one-to-one mapping information. The TP then obtains measurement information for the identified UE, enabling better control and communications for uplink and downlink transmissions between multiple TPs and the UE.

Abstract: A service deployment method and a network functions acceleration platform are provided. The method includes: when an FPGA is powered on, loading, by the FPGA, an FPGA framework file, so that the FPGA includes M partial reconfigurable PR areas, a configuration module, and a data flow forwarding module; when receiving a service configuration instruction, generating, by a physical machine, a VM, and selecting at least one PR area from the M PR areas to establish a correspondence with the VM; performing, by a configuration module, service resource configuration on the selected PR area according to a PR configuration resource, so that the selected PR area has a service processing capability; adding, by the configuration module, a forwarding entry corresponding to the PR area in a forwarding flow table, so that the data flow forwarding module forwards a received network packet to the corresponding PR area.

Abstract: Embodiments of the present invention provide a data transmission control node, a system, and a data transmission management method. A TCP subflow on which data offloading needs to be performed is selected according to a requirement for running the system. By changing an input item of an MPTCP congestion control algorithm, data on the selected TCP subflow can be migrated to another expected TCP subflow that belongs to a same MPTCP connection as the selected TCP subflow. By taking an additional measure to manage and control MPTCP data transmission on an access network side, a result of the data offloading on the TCP subflow can satisfy the expected requirement for running the system, so as to optimize system performance, and improve practicability of the MPTCP congestion control algorithm.

Abstract: This application relates to the field of communications technologies, and discloses example polar coding methods and apparatus, so as to improve accuracy of a reliability order of polarized channels. One example method includes determining a sequence used to code to-be-coded bits, and performing polar coding on the to-be-coded bits by using the sequence to obtain coded bits, where the sequence is used to represent a reliability order of N polarized channels, where N is a mother code length of a polar code, and wherein N is a positive integer power of 2.

Abstract: Relating to the field of video coding, a method and an apparatus for building a motion vector list for motion vector prediction, which solve a problem of building motion vector lists of at least two PUs in a same coding unit in a serial manner, and improve the parallel processing capability. The method includes: obtaining spatial neighboring blocks of a current prediction block, where the current prediction block is located inside a current coding unit; determining available neighboring blocks of the current prediction block according to a partition manner of the current coding unit, where the available neighboring blocks are located outside the current coding unit; and obtaining motion vectors from the available neighboring blocks in a preset sequence according to motion vectors of the available neighboring blocks, and adding the obtained motion vectors to the motion vector list.

Abstract: Embodiments of the present invention provide a channel-state information process processing method, a network device, and a user equipment, where the channel-state information process processing method includes: after receiving a first channel-state information CSI request sent by a first network device, if CSI corresponding to multiple aperiodic CSI processes has not been reported by a user equipment, dropping CSI corresponding to a part of aperiodic CSI processes among the multiple aperiodic CSI processes, where each CSI process is associated with a channel measurement resource and an interference measurement resource. A problem existing after a CoMP technology is introduced can be solved that the UE cannot implement processing of multiple CSI processes.

Abstract: The present invention relates to the field of communications technologies and discloses an antenna system, and the antenna system includes: a radiating element, at least one strip line ground plane and signal cavity, and at least one inner conductor, where the radiating element includes radiation arms, radiation baluns, and feeding inner cores; the strip line ground plane is used as a reflective surface of the radiating element; the baluns of the radiating element are electrically connected to the strip line ground plane, and all of at least two feeding inner cores of the radiating element are electrically connected to one inner conductor. Therefore, a feeding manner of an existing array antenna can be optimized very conveniently, assembly time is greatly reduced, quantities of welding points and cables are reduced, and consistency and reliability are improved.

Abstract: The present disclosure provides a method for encoding an intra-frame prediction mode, including: obtaining an intra-frame prediction mode of a current intra-frame encoding block from a preset prediction mode set; obtaining reference prediction modes of the current intra-frame encoding block, where the reference prediction modes are intra-frame prediction modes of available adjacent blocks of the current intra-frame encoding block or prediction modes in a preset backup reference mode set; writing a first flag bit into a code stream according to the reference prediction modes and the intra-frame prediction mode; and, when the intra-frame prediction mode of the encoding block is different from all the reference prediction modes, obtaining a prediction mode encoding value according to a size relationship between the value of the intra-frame prediction mode and values of the reference prediction modes, and encoding the prediction mode encoding value.