Abstract: A data synchronization device is provided, including a control module, a data block storage module and an encoding/decoding module. A data synchronization method using the device is also provided, including: selecting, by the control module, a data transmission network node and selecting the to-be-synchronized data block of a data transmission network node from the data storage module; encapsulating the to-be-synchronized data block in a specific format and then broadcasting the encapsulated data, where if there are two or more to-be-synchronized data blocks, the to-be-synchronized data blocks are encoded and then encapsulated; otherwise, the to-be-synchronized data block is directly encapsulated; processing, by the encoding/decoding module, the encapsulated data to obtain the to-be-synchronized data block, and finally inputting the obtained to-be-synchronized data block into the data block storage module.

Abstract: A deep reinforcement learning (DRL)-based intelligent job batching method and apparatus, and an electronic device are provided. The method includes: obtaining static features and a dynamic feature of each job, where the static features of the job include a delivery date, a specification and a process requirement of the job, and the dynamic feature of the job includes a receiving moment; and inputting the static features and the dynamic feature of each job into a job batching module, and using a Markov decision process (MDP) by the job batching module to combine jobs with similar features in a to-be-batched job set into an identical batch, so as to minimize a total quantity of batches obtained finally and a difference in features of jobs in each batch. The DRL-based intelligent job batching method and apparatus can learn a stable batching strategy and provide a stable and efficient job batching solution.

Abstract: Embodiments of the present disclosure provide a software defined network publish-subscribe system and method. The software defined network publish-subscribe system comprises a controller, an administrator controller, a sequencing subsystem and a message multi-backup cache subsystem. The controller is configured to process publish-subscribe transactions in clusters, and comprises sequencing nodes used for generating ordered serial numbers for messages and cache nodes for caching the messages. The administrator controller is configured to simultaneously perform functions of the controller and manage software-defined network. The sequencing subsystem is configured to realize the cooperative sequencing between the administrator controller and the sequencing nodes. The message multi-backup cache subsystem is configured to realize the consistency of the messages through the administrator controller and the cache nodes.

Abstract: Disclosed are a method and device for implementing an ad hoc network routing protocol in a multi-agent system. The device includes a trajectory information acquisition module, a data packet delivery module, a delivery confirmation module, a data packet forwarding module, a data transmission feedback module and a data storage module. The method includes: acquiring a separation and rendezvous timing table and an adjacency matrix of an ad hoc network composed of multiple agents at a given moment; if a source node has message sending requirement, performing data packet delivery and delivery confirmation based on the separation and rendezvous timing table and the adjacency matrix; and performing data packet forwarding and transmission status feedback by a non-source node.

Abstract: Embodiments of the present disclosure provide an Internet-of-Things resource access system and method. The system comprises a protocol management subsystem, a data conversion subsystem, and a load balancing subsystem.

Abstract: Disclosed is a method for network slicing, which includes: determining a service keyword of a service intention after receiving the service intention from a user; converting the service intention into a network request according to the service keyword of the service intention on the basis of a pre-established mapping model, wherein, the network request comprises at least one network requirement index; determining a set of network slicing configuration parameters according to the network request and current network environment state on the basis of a network slicing model, wherein the network slicing model is a deep neural network model for determining network slicing configuration parameters according to a network request and a network environment state; and performing a network slicing according to the set of network slicing configuration parameters. The present disclosure also provides a device for network slicing and a computer readable storage medium.

Abstract: An information processing method, and the method includes: receiving, by a base station, a buffer status report BSR sent by a first user equipment UE; sending, by the base station, scheduling information to a corresponding second UE based on at least one of a number of the first UEs and the BSR; and allocating SL resource to the first UE according to received sidelink SL resource perception information reported by the second UE based on the scheduling information.

Abstract: Disclosed is a data sharing method, including: building a trust alliance block chain, the trust alliance block chain comprising a main chain and at least one slave chain; each slave chain corresponding to a domain; and each domain comprising a leader node; establishing a virtual slave chain on the trust alliance block chain; designating, by the leader node in each domain, a node as a federated learning node; assigning, by the leader node in each domain, a virtual identity to the federated learning node to join the federated learning node to the virtual slave chain; and performing, by each federated learning node, a joint training on a local federated learning model using data generated in its own domain to establish a public federated learning model, through which data are shared among the domains. The present disclosure further provides a data sharing system, an electronic device and a storage medium.

Abstract: A frame transmission method, includes: sending, by a first communication end to a second communication end, a first RTS on an unlicensed channel, wherein the first RTS carries a licensed frequency band source indication; receiving, according to the licensed frequency band source indication, an assist-clear to send (A-CTS) that is sent by the second communication end in a retreat state on the licensed channel; and sending, according to the A-CTS, a first clear-request to send (C-RTS) on the unlicensed channel, wherein the first C-RTS is configured for instructing a third communication end to exit the retreat state entered based on the first RTS.

Abstract: A method for transmitting transmission block configuration parameter is provided. The method includes: sending first downlink control information (DCI) for scheduling a multi-transmission time interval transmission block, where the first DCI includes a preset indicator indicating that a configuration parameter that determines a transmission block size (TBS) of a retransmission TB scheduled by means of the first DCI is from the first DCI or second DCI, and the second DCI is DCI for scheduling a new transmission TB corresponding to the retransmission TB.

Abstract: Embodiments of the present invention provide a method and apparatus for accelerating distributed training of a deep neural network. The method comprises: based on parallel training, the training of deep neural network is designed as a distributed training mode. A deep neural network to be trained is divided into multiple sub-networks. A set of training samples is divided into multiple subsets of samples. The training of the deep neural network to be trained is performed with the multiple subsets of samples based on a distributed cluster architecture and a preset scheduling method. The multiple sub-networks are simultaneously trained so as to fulfill the distributed training of the deep neural network. The utilization of the distributed cluster architecture and the preset scheduling method may reduce, through data localization, the effect of network delay on the sub-networks under distributed training; adapt the training strategy in real time; and synchronize the sub-networks trained in parallel.

Abstract: Disclosed is a method of network resource allocation. The method includes: generating an adjacency matrix of nodes in a metropolitan area network (MAN) according to spatial adjacency relationships of the nodes; generating a network state feature matrix according to traffic information of each node; extracting traffic spatial features of the nodes from the adjacency matrix and the network state feature matrix through a traffic spatial feature extraction model; obtaining predicted traffic of the nodes from the traffic spatial features through a traffic prediction model; and performing a network resource allocation according to the predicted traffic of the nodes. Further, a device of network resource allocation and a non-transitory computer-readable storage medium are also disclosed.

Abstract: The present disclosure provides a fine-grained image recognition method, an electronic device and a computer readable storage medium. The method comprises the steps of feature extraction, calculation of feature discriminant loss function, calculation of feature diversity loss function and calculation of model optimization loss function. The present disclosure comprehensively considers influences of factors such as a large intra-class difference, a small inter-class difference, and a great influence of background noise of the fine-grained image, and makes constrains such that the feature maps belonging to each class are discriminative and have the features of corresponding class, thus reducing the intra-class difference, decreasing the learning difficulty and learning better discriminative features.

Abstract: An antenna configured to transmit or receive a signal is provided. The antenna includes a substantially transparent base substrate and an antenna electrode on the substantially transparent base substrate. The antenna electrode includes a substantially transparent conductive layer, and a first conductive line abutting an edge portion of the substantially transparent conductive layer and electrically connected to the edge portion of the substantially transparent conductive layer. An electrical conductivity of the first conductive line is greater than an electrical conductivity of the substantially transparent conductive layer.

Abstract: Disclosed is a hybrid automatic repeat request (HARD) feedback enhancement method. The method comprises: sending downlink control information comprising first indication information to a terminal, the first indication information being used for indicating the number of downlink transport blocks received by the terminal on a sub-frame after the current sub-frame and a sub-carrier after the current sub-carrier within the current HARO feedback window.

Abstract: Examples of the present disclosure provide a method for setting up a route path, a device and a computer-readable storage medium. The method may include: establishing a neighbor list; wherein, the neighbor list comprises neighbor nodes of the node and position information of the neighbor nodes; in response to receiving a routing request (RREQ) message from a previous hop node, recording a source Internet Protocol (IP) address of a source node contained in the RREQ message as a destination IP address in a routing table established in the node; recording the previous hop node as a next hop node corresponding to the destination IP address in the routing table; selecting a pre-determined number of nodes from the neighbor list as next hop nodes of the RREQ message according to the position information of the neighbor nodes; and sending the RREQ message to the next hop nodes.

Abstract: A data transmission method includes: sending, by a network device, a fixed-length downlink control information (DCI), wherein the fixed-length DCI comprises a detection parameter corresponding to a variable-length DCI.

Abstract: The embodiments of the present disclosure provide data transmission methods and apparatuses, a communication device and a storage medium. The data transmission method includes: receiving transmission feedback for one or more Code Block Groups (CBGs); transmitting one or more transmission blocks (TBs) based on the transmission feedback, where one of the TBs includes one or more retransmission CBGs and one or more initial transmission CBGs.

Abstract: Disclosed are a method and an apparatus for handover between heterogeneous networks.

Abstract: The invention discloses a cross-layer network fault recovery system and method based on configuration migration.