Abstract: The present disclosure provides an annular anchor, a method for calculating antitorque bearing strength thereof and an installing and recycling assemble for the anchor, relating to the technical field of ocean engineering. The annular anchor includes an anchor body opened at its upper and lower ends, and a connection portion arranged on the outer side wall of said anchor body for connecting with parts to be moored. The annular anchor provided by the present disclosure omits the top cover as the structure prevailing in the traditional suction anchor, avoiding the drawback of installing the traditional suction anchor on the seabed, instead the annular anchor can be installed to a certain depth below the seabed. During the descent, there is no need to take into account the resistance caused by the top cover, so that the annular anchor can be installed simply and quickly.

Abstract: Disclosed are a method and apparatus for task scheduling based on deep reinforcement learning and a device. The method comprises: obtaining multiple target subtasks to be scheduled; building target state data corresponding to the multiple target subtasks, wherein the target state data comprises a first set, a second set, a third set, and a fourth set; inputting the target state data into a pre-trained task scheduling model, to obtain a scheduling result of each target subtask; wherein, the scheduling result of each target subtask comprises a probability that the target subtask is scheduled to each target node; for each target subtask, determining a target node to which the target subtask is to be scheduled based on the scheduling result of the target subtask, and scheduling the target subtask to the determined target node.

Abstract: The present invention provides a method for optimizing the energy efficiency of wireless sensor network based on the assistance of unmanned aerial vehicle, firstly, collecting the state of the WSN through current routing scheme, and inputting the state of the WSN into the decision network of the agent to determine a next hover node; Secondly, based on the location of the next hover node, generating a new routing scheme by the UAV, and sending each sensor node's routing to its corresponding sensor node through current routing by the UAV; Lastly, after all sensor nodes have received their routings respectively, all sensor nodes send their collected data to the hover node through their routings respectively, and the UAV flies to and hovers above the next hover node to collect data through the next hover node, thus the data collection of the whole WSN is completed.

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: A network resource scheduling method, apparatus, an electronic device and a storage medium are disclosed. An embodiment of the method includes: upon receipt of a network data stream, determining a traffic type of the network data stream based on the number of data packets of the network data stream received within a specified period of time, lengths of the data packets and reception times of the data packets; for each data packet comprised in the network data stream, determining a target transmission path for the data packet, based on node state parameters of nodes in the network cluster, link state parameters of links in the network cluster, and the traffic type of the network data stream when the data packet is received; and transmitting the data packet via the target transmission path.

Abstract: Disclosed are a method and apparatus for task scheduling based on deep reinforcement learning and a device. The method comprises: obtaining multiple target subtasks to be scheduled; building target state data corresponding to the multiple target subtasks, wherein the target state data comprises a first set, a second set, a third set, and a fourth set; inputting the target state data into a pre-trained task scheduling model, to obtain a scheduling result of each target subtask; wherein, the scheduling result of each target subtask comprises a probability that the target subtask is scheduled to each target node; for each target subtask, determining a target node to which the target subtask is to be scheduled based on the scheduling result of the target subtask, and scheduling the target subtask to the determined target node.

Abstract: A network resource scheduling method, apparatus, an electronic device and a storage medium are disclosed. An embodiment of the method includes: upon receipt of a network data stream, determining a traffic type of the network data stream based on the number of data packets of the network data stream received within a specified period of time, lengths of the data packets and reception times of the data packets; for each data packet comprised in the network data stream, determining a target transmission path for the data packet, based on node state parameters of nodes in the network cluster, link state parameters of links in the network cluster, and the traffic type of the network data stream when the data packet is received; and transmitting the data packet via the target transmission path.

Abstract: Embodiments of the present invention provide a congestion control method and apparatus based on a software defined network SDN, and an SDN controller. The method comprises: obtaining a packet_in message sent by a switch; determining a data packet included in the packet_in message; performing a first congestion control processing for a network where the SDN controller is located based on a topological structure and link information of the network when the data packet is a handshake information SYN packet for requesting to establish a TCP connection; performing a second congestion control processing for the network based on the link information when the data packet is a finish information FIN packet for responding to disconnection of a TCP connection; deleting information of a TCP connection stored in a database and corresponding to the data packet when the data packet is an FIN packet requesting to disconnect a TCP connection.

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: Embodiments of the present invention provide a congestion control method and apparatus based on a software defined network SDN, and an SDN controller. The method comprises: obtaining a packet_in message sent by a switch; determining a data packet included in the packet_in message; performing a first congestion control processing for a network where the SDN controller is located based on a topological structure and link information of the network when the data packet is a handshake information SYN packet for requesting to establish a TCP connection; performing a second congestion control processing for the network based on the link information when the data packet is a finish information FIN packet for responding to disconnection of a TCP connection; deleting information of a TCP connection stored in a database and corresponding to the data packet when the data packet is an FIN packet requesting to disconnect a TCP connection.

Abstract: A method for selecting and configuring network supernodes includes (i) in a first set period T1, each node, other than a control node in a network, regularly sending information on interactions between the node and other nodes to the control node; and (ii) in a second set period T2, the control node dividing the nodes into a plurality of node clusters according to the received information on interactions among the nodes and selecting supernodes from each node cluster, where each node belonging uniquely to a node cluster.

Abstract: A method for selecting and configuring network supernodes including the following operational steps: in a first set period T1, each node other than the control node in a network sends information on interactions between the node and other nodes to the control node regularly; and in a second set period T2, the control node divides the nodes into a plurality of node clusters according to the received information on interactions among the nodes, and selects supernodes from each node cluster, each node belonging uniquely to a node cluster. The supernode selection according to the invention takes into account both performance and resource conditions of itself and interactions with other nodes. Therefore, when implementing control over other nodes in the node cluster to which it belongs, the selected supernode can find the corresponding node in a short time and shorten the searching time and path, thereby improving the working efficiency.