46results about How to "Improve packet loss rate" patented technology

The invention discloses a controlling method for a burst assembly, a corresponding device and a communication equipment thereof. A core of the method and the device in the invention includes the following steps: by comparing an assembly time with a time threshold set in advance, a change trend of carriers in a buffer queue is judged; if the change of the carrier burst is increased or reduced, a length threshold is directly increased or reduced; if the carriers are in an increasing trend, the threshold is randomly increased according to a certain probability; if the carriers are in a reducing trend, the threshold is randomly reduced according to a certain probability; if the carrier is normal, the length threshold is not adjusted, therefore, the length threshold can realize dynamic adjustment with the changes of the carrier. By adopting the method and the device provided by the invention, a delay from a business end to anther business end is assured, thus optimizing the performance of the network.

The utility model discloses a network data processing method and device; wherein, the method comprises: a plurality of input ends are configured with priorities; input data from the input ends with priorities or high priorities is exchanged and processed first according to the priorities of the input ends; data exchanged and processed is exported; the network device comprises: a priority configuration module, which configures priorities or high priorities to the input ends with smaller bandwidth or more important input ends; a data exchange module, which exchanges and processes input data from the input ends with priorities first; an output module, which exports exchange data. The utility model is applicable to the network device with important input ends to improve the transmission speed and quality of data from the important input ends.

The invention relates to an Aloha enhanced access method for a low-orbit constellation system, belongs to the field of satellite communication, and is used for solving the problem of quick access whena high-dynamic terminal communicates with a new generation of low-orbit constellation system. The method comprises the following steps: firstly, constructing a system satellite access system model; reserving different numbers of channels for users with different priorities of the related high-dynamic terminals in advance; when the user terminal has an access demand, calculating the specific moment of sending the data packet by utilizing access control; when the data packet is sent, enabling the user terminal to divide a channel into time slots by utilizing a local clock and randomly puts thedata packet into the time slots to be sent; finally, utilizing the characteristic that the user terminal is covered by multiple satellites of the system in the communication process, wherein after receiving the data packet, each receiving end satellite forwards the data packet to the low-orbit constellation system for interference elimination until all the data packets are successfully decoded orthe maximum number of iterations is reached. According to the method disclosed in the invention, the throughput rate and packet loss rate indexes of the system can be optimized, the resource utilization rate is improved, and the first access success rate of the high-dynamic terminal can be optimized.

The invention discloses a load balancing method based on a graph convolutional neural network and deep reinforcement learning. The load balancing algorithm based on the graph convolutional neural network and deep reinforcement learning comprises the following steps: acquiring a network state undirected graph, wherein the network state undirected graph comprises a characteristic matrix of switches and data link load weight information between the switches; obtaining stream information; obtaining a trained DQN decision model; and inputting the flow information and the network state undirected graph into the DQN decision model to obtain a decision action. According to the method, deep reinforcement learning and the graph convolutional neural network are combined and applied to a load balancing algorithm, so that the model can make a decision according to state information, the topological structure of the network is considered as a decision factor, the model can make a decision according to a more comprehensive network state, and the decision performance of the model is improved.

A method for optimizing OBS / OPS network performance based on virtual node includes combining a numbers of physical nodes on service shortest route to be one virtual node according to network service flow rate of last statistic time section , setting special wavelength to connect head and tail physical node in virtual node to let data packet jump over at least two physical nodes with only one time of processing when data packet is via this virtual node for lowering end to end time delay and packet lost rate in network .

Disclosed are 5G core network equipment and method supporting PDN link optimization, which relate to the technical field of mobile communications, are implemented on the basis of an AMF entity, a UDM entity, a PCF entity, an SMF entity, a UPF entity and a remote PDN gateway. The specific process comprises: a UE sending a session establishment request carrying DNN information to the AMF entity; the AMF entity selecting an SMF entity according to the request; the SMF entity initiating a session policy request and requesting session policy information of a target session from a PCF entity; the PCF entity issuing session policy information to the SMF entity; the SMF entity judging whether the requested DNN information supports an N6 interface MPTCP proxy or not, if yes, the SMF entity continuing to judge whether the session policy information carries a correct MPTCP proxy working mode and priority or not, and initiates a session establishment or modification process carrying an MAR rule cell to the UPF entity; the UPF entity analyzing the MAR rule cell and configuring an MPTCP proxy entity function; and after the session is established, the UPF entity successfully establishing MPTCP proxy connection with the remote PDN gateway. According to the invention, a plurality of PDN link quality optimization functions can be realized.

The invention discloses a cross-layer vehicular network routing method based on forwarding efficiency prediction. According to the method, through combination of an existing MAC layer access protocol, by taking transmission failure probability resulting from channel fading into consideration, an effective transmission distance of nodes in single-time transmission is defined. Moreover, through combination of time slot distribution conditions of multi-channel MAC layers access on a control channel and the data processing capabilities of the nodes, the waiting time delay of single-time transmission is defined. A forwarding frequency of single-time forwarding is defined by using the two parameters. Through prediction of forwarding efficiencies of all selectable nodes in a next hop and routing selection by taking unit time transmission distance optimization as a strategy, forwarding can be carried out by using the nodes more effectively. The method aims at improving the effectiveness of each time of forwarding; production of the low efficiency forwarding is reduced; the average end to end time delay is reduced; and a packet loss rate is reduced.

The invention discloses a controlling method for a burst assembly, a corresponding device and a communication equipment thereof. A core of the method and the device in the invention includes the following steps: by comparing an assembly time with a time threshold set in advance, a change trend of carriers in a buffer queue is judged; if the change of the carrier burst is increased or reduced, a length threshold is directly increased or reduced; if the carriers are in an increasing trend, the threshold is randomly increased according to a certain probability; if the carriers are in a reducing trend, the threshold is randomly reduced according to a certain probability; if the carrier is normal, the length threshold is not adjusted, therefore, the length threshold can realize dynamic adjustment with the changes of the carrier. By adopting the method and the device provided by the invention, a delay from a business end to anther business end is assured, thus optimizing the performance of the network.

The embodiment of the invention discloses an internetwork interference avoiding method based on a nested composite superframe structure. The method comprises the nested composite superframe structure, improved frame structures and a communication mechanism under the composite superframe. In adoption of the embodiment of the invention, the internetwork interference can be effectively avoided; and performances such as the packet loss probability and the throughput capacity can be improved.

The invention relates to an interference sensing cross-layer routing method based on node sending and receiving capacity. The method defines parameters for judging the sending and receiving capacity of a node based on an 802.11 medium access control (MAC) layer, and the two parameters can reflect that strength and weakness of interference of the periphery of the node and communication load degree born by the node, based on the fact, a MAC layer and a network layer are combined and a new routing criterion is defined, the new criterion is related to the number of surrounding active neighbor nodes, the number of data flow passing the nodes, capacity of node access channel and topology changing conditions around the nodes, and real conditions of the network can be further reflected. The interference sensing cross-layer routing method based on the node sending and receiving capacity aims to reduce interference, selects a path which is largest in bottleneck node comprehensive capacity or a path which is smallest in bottleneck link interference as far as possible, and the path which is largest in the bottleneck node comprehensive capacity or the path which is smallest in the bottleneck link interference is used as routing, and therefore average end-to-end delay and packet loss probability are reduced, and the throughput is improved.

The invention discloses a supporting node mobility method based on node speed, a new neighbor node is obtained through judging whether a mobile synchronization frame MSYNC which is sent by a mobile node is replied or not by the method, the distance between the new neighbor node and the mobile node is obtained by the new neighbor node by receiving the strength of a MSYNC signal, meanwhile, whetherreplying MACK or not to the mobile node is determined through a judgment condition; after the judgment condition is met, the time interval for sending the MSYNC next time by the mobile node is calculated by the new neighbor node through the static neighbor information of the new neighbor node and the state information of the mobile node. The judgment condition is that: an included angle of a directed line segment and the moving direction of a mobile node m is bigger than 90 degrees, and the distance between the mobile node m and a new neighbor node n is less than a transmission range R. The energy consumption which is caused by frequent periodicity neighbor searching is avoided by the method in the invention, the scheduling information of a new cluster is quickly obtained when the mobile node moves among clusters, and mass lost packages which are caused by the differences of period scheduling are reduced.

The present invention is based on the dynamic routing method of the three-dimensional heterogeneous network in the emergency scene of the UAV. The method can adapt to the fusion network of dynamic topology changes, and the UAV is used as the relay node to form the relay node, and the nodes in the same area do not communicate. , the node communication in different areas is routed through the UAV node relay, and the ground node is only used as the end node in the route. In each communication process, the method includes the following steps: Step A: respectively Nodes model the scene; step B: improve the link weight on the basis of the traditional OLSR protocol, and select the routing path as cost=α×ETT+β×E i +(1‑α‑β)×T as the benchmark; Step C: The core idea of ​​the OLSR routing protocol is to select the MPR set to reduce routing overhead. The improved LBMRE‑OLSR abandons the classic MPR set selection algorithm, The selection of the MPR set is based on the node's expected transmission time, node energy and node congestion degree.

An adaptive service data distribution method based on opportunistic communication in the Internet of Vehicles environment. Adjacent vehicles or RSUs receive Beacon messages to judge the validity of the messages. If the forwarding conditions are met, the expected forwarding delay is calculated and the expected forwarding delay value is added to the field. Return RREP. After the last hop vehicle or RSU receives the RREP from its adjacent object, it executes the RREP response processing process and returns RREQ to the optimal next hop selected from the candidate object set. After receiving the RREQ, the candidate object judges whether All data packets arrive at the target location, if not all reach the target location, the forwarding vehicle or RSU will restart a new round of broadcasting until all data packets reach the target vehicle. The invention realizes the effective distribution of service data in the Internet of Vehicles environment, and improves the packet loss rate, reduces the number of retransmissions, delay and network overhead and other network performance under the condition of ensuring that other indicators are not damaged.

The present invention provides a cross-layer vehicle network routing method based on a link transmission capacity. In a vehicle network city scene, the method combines an existing MAC layer access protocol to reflect transmission effectiveness between nodes through defining and representing parameters of the communication capacity between the nodes through considering the transmission failure probability generated by channel fading and neighbor node channel usage condition; and on this basis, a new routing metric suitable for city scenes is given to realize the protocol design of a network layer and an MAC layer so as to more effectively utilize the nodes for forwarding. The cross-layer vehicle network routing method based on the link transmission capacity takes the improvement of the link stability as an objective and reduces the number of times of repeatedly establishing a route as far as possible so as to reduce the average end-to-end time delay and reduce the packet loss probability.