33results about How to "Improve data forwarding efficiency" patented technology

The invention discloses a method and a system for mapping HTTP service to an information center network. The method comprises steps that: a user and a server are respectively connected with an inlet end and an outlet end of the information center network, an HTTP request sent by the user is parsed at the inlet end, and the HTTP request is assembled into an information center network request message which is then sent to the outlet end through a router in the information center network; the request message is assembled by the outlet end into HTTP information, and data is requested from the server; the HTTP response information and TCP data information which are returned from the server are received by the outlet end, after the information is assembled into the information center network data message, and the information center network data message is returned to the inlet end through the router of the information center network; the information center network data message returned from the outlet end is received by the inlet end, and the information center network data message is assembled into HTTP response data or HTTP data which are returned to the user. The method solves a problem of compatibility between a network in the prior art and the information center network, and increment type disposition of the information center network can be realized.

A data forwarding method for realizing local area network virtual channel transmission, comprising the following steps: first judging whether a data packet arrives at an interface of a device, and checking the frame check code of the data packet, and secondly judging whether the type of the incoming data packet matches the type of the input interface , and check whether the "label forwarding information data table" clearly indicates the output interface of the data packet or the VLAN ID of the output data packet according to the input interface or input data packet, and the device completes according to the instructions of the "label forwarding information data table". When data packets are exchanged between any two interfaces of the device, the virtual local area network identification VLAN ID is exchanged according to the specified. The present invention breaks through the limitations of LAN networking caused by the lack of VLAN IDs stipulated by IEEE802.1Q, and realizes the end-to-end interconnection of large-scale VLANs at low cost without using new communication protocols under the existing relevant protocols of the LAN. , it can not only support centralized broadband LAN access, but also provide LAN VPN and PPPOE access services at the same time, and network operators do not need to increase the cost of learning new technologies.

The invention discloses a method for predicting the link quality of a wireless sensor network, which includes the following steps: S1, using a Kalman filter algorithm to reduce the noise of the link quality parameters, using the maximum and minimum method to standardize the link quality parameters, and using The fuzzy C-means clustering algorithm divides the link quality level; S2, uses the XGBoost algorithm to train the link quality prediction model, and predicts the link quality level at the next moment through the prediction model. Considering the dynamic change of wireless sensor network link quality, the present invention provides an XGBoost-based wireless sensor network link quality prediction method, which can effectively predict the link quality at the next moment, and can improve network throughput, thereby improving The data forwarding efficiency of the network is improved, and it provides a basis for the upper layer routing protocol to select the communication link. In addition, this method has the advantages of prolonging the life of the network and saving node energy.

The invention discloses a community-type opportunistic network data forwarding method. In the source node community, data packets are forwarded in a single-copy forwarding manner. By calculating the central utility value of the neighbor node, the data packet is sent to the neighbor node in which the central utility value is the largest. node, repeat this process until it is sent to the destination node. When the data packet is forwarded between communities, because the community where the destination node is located is unknown, it is impossible to know exactly which community it should be sent directly to, so the multi-copy forwarding method is used to increase the delivery rate, and the data packet is sent to the drop boxes of other communities , and query the node set N corresponding to the community a to find the destination node, and then forward the data packet to the destination node. The method of the invention is directly applied to the community opportunity network, so that the network transmission speed becomes faster and the message load becomes smaller.

The invention discloses a network processor and a data processing method thereof; the network processor comprises a driver module, a statistics module and a micro-code module; the data processing comprises an analysis phase for carrying out analysis to a data packet; the data processing method includes the following steps: A. the driver module writes down the configuration information of at leastone between the port of the network processor and the VLAN into the statistics module; B. in the analysis phase, the micro-code module accesses the statistics module and then obtains the configuration information, and later the searching of key values is carried out according to the configuration information structure. By writing down the configuration information of the port and / or the VLAN intothe statistics module, the micro-code module uses the configuration information of the statistics module structure to construct the searching key value, thereby avoiding constructing redundant searching key values for the business and function without configuration in the Parse phase, saving the instructions and enhancing the data transmitting efficiency of the network processor.

The invention discloses an Nginx upstream proxy service system and an implementation method, the proxy service system includes a management plane proxy service module and a data plane Nginx module, and the method includes: performing the management plane proxy service module and the data plane The Nginx module establishes a shared memory; the shared memory is used as a bearer channel of the TCP connection load data; the TCP connection load data includes the user's HTTP request data and the HTTP response data of the upstream server. The system and method in the present invention effectively improve the data forwarding efficiency of the proxy system by using the shared memory as the bearer channel of the TCP connection load data.

The invention relates to a heuristic routing method utilizing position information in a shipborne adhoc network. The heuristic routing method is characterized by comprising three main modules, namely a geographical position based routing strategy, a topologic void crossing method and a topologic void prediction model, and two auxiliary modules, namely a geographical position service model and a wireless network constraint strategy. According to the heuristic routing method, the existence of the topologic void can be predicted in advance, and the topologic void can be crossed by a geographical information based heuristic algorithm, so that the core technical problem of high-efficiency data forwarding in the shipborne adhoc network is solved.

The utility model relates to a method for data forward, which is suitable for the message transmission between STA. The method is characterized in comprising a forward table for recording the corresponding relationship of STA and AP is arranged on the AP; the source AP can search whether the AP corresponding to the target STA is existed in self forward table after receiving the message emitted tothe target STA by the source STA; if existed, the received message can be forwarded directly to the AP corresponding to the target STA for processing; if not existed, the received message can be forwarded to AC for processing. In addition, the utility model also provides an access point device. With the technical proposal, the utility model can lighten the AC processing burden and reduce the AC performance requirement.

The invention provides a PMIPv6 switching method based on a busy forward chain. By the adoption of the PMIPv6 switching method based on the busy forward chain, unnecessary signaling overheads can be avoided, rapid MN switching is guaranteed, and the mobile management performance of PMIPv6 can be optimized. The PMIPv6 switching method based on the busy forward chain comprises the following steps that (1) switching is triggered, wherein a switching triggering message is sent to a current MAG to which an MN has access; (2) a switching decision is made, wherein the switching decision is made by the current MAG according to the switching triggering message and a threshold K; (3) switching is executed, wherein the forward chain is lengthened by the current MAG to a target MAG if the MN is in the idle state and the length of the forward chain is smaller than the threshold K, rapid switching is initiated if the MN is in the idle state and the length of the forward chain is equal to or larger than the threshold or the MN is in the busy state, and the target MAG is informed that the forward chain is in a loop state by the current MAG through a PBU message when the forward chain is in the loop state, so that the target MAG directly executes caching and waits for MN access instead of conducting forwarding according to the forward chain when the target MAG receives an MN data packet.

The invention discloses data transmission method when the route region is refreshed. The method comprises following steps: MS requires the new SGSN of new route region to attain the PDP context from the old SGSN of old route region when the route region is refreshed in SGSN; the new SGSN arranges the tunnel to each PDP context after receiving the PDP context from old SGSN and arranges the IP address of relative PDP context to each tunnel, then, sending the relation between distributed tunnels and IP addresses to the old SGSN; the old SGSN transmits the data of each PDP context to the new SGSN via relative tunnel. The inventive method solves the problem that old SGSN can only transmit the data to unique address in present technique, so the new SGSN should relates to distribution modules for distributing data. The invention can make the transmission of user's data more flexible when the route region of SGSN is refreshed, simplify the design of system and shorten the refresh time of route region.

The disclosed device includes following parts: arithmetic unit including microprocessor in use for processing data and controlling circuit in realtime; Ethernet communication unit for sending and receiving data to / from Ethernet under control of arithmetic unit; multi serial ports communication unit including chip of multi serial ports and logic control circuit in use for data interaction with I / Ocards under control of arithmetic unit; the logic control circuit for receiving control signals from the arithmetic unit and chip of multi serial ports, for switching I / O channels in chip, and for controlling serial ports to transmit / receive data; power circuit. Features are: raised efficiency for forwarding data, network scale, reliability and response speed.

The present application discloses a method, device, storage medium, and electronic device for aggregation and distribution based on a programmable switching chip. The comparison result determines whether the parameters corresponding to the information to be matched need advanced filtering; when the parameters corresponding to the information to be matched need advanced filtering, obtain preset custom parameters and insert them into the first packet to generate a second packet Send from channel to compute unit. The calculation unit receives and analyzes the second packet, extracts parameters, detects and generates a detection result; judges whether the second packet needs advanced filtering based on the detection result. Therefore, by adopting the embodiment of the present application, the programmable switching chip can make full use of the programmability of the "programmable switching unit", which can effectively release the computing power and memory of the "computing unit", reduce forwarding delay, improve forwarding efficiency and overall machine throughput Rate.

A method for predicting the link quality of a wireless sensor network using GRU, the method comprising: obtaining a link quality parameter of the wireless sensor network at each moment within a period of time, and determining the link quality parameter of each moment according to the obtained link quality parameter Quality level; calculate the Pearson correlation coefficient between the historical link quality parameter before the current moment and the link quality parameter at the current moment; determine the quantity of the historical moment with the current moment Pearson correlation coefficient greater than the threshold value, and according to the The above quantity determines the size of the time window, and the link quality parameters are intercepted with the time window of the size to obtain the training sample set, and the training sample set with the link quality level label is applied to the GRU neural network model Training: using the trained GRU neural network model to predict the link quality of the wireless sensor. The invention can effectively and accurately predict the link quality at the next moment.