158results about How to "Save CPU resources" patented technology

The invention discloses a method and a device for transmitting message. The method is as follows: a virtual switch receives the first flow stable and a second flow stable sent by the SDN controller; if the message is matched with the first flow table while the virtual switch receives the message from the virtual machine, the virtual switch sends the message to the virtual NAT device via the first flow table, the virtual NAT device does NAT conversion on the source IP address of the message and sends the conversion NAT message to the virtual switch; if the message is matched with the second flow table while the virtual switch receives the message from the virtual NAT device, the virtual switch uses the second flow table for executing VXLAN packaging operation on the message and sends the VXLAN packaged message to the VXLAN gateway. According to the embodiment, the NAT conversion is avoided to the message of the virtual machine by the VXLAN gateway, the CPU resource of the VXLAN gateway is saved and the processing performance of the VXLAN gateway is reduced.

The invention discloses a handwriting generating method for an electronic writing brush, which relates to the technical field of computer image processing and includes the following steps: S1 calculating distance between two adjacent touch points according to touch point coordinates on a touch track delivered by a touch device; S2 if the distance is smaller than a preset threshold, the touch point is obtained according to the preset threshold serving as the distance, and if the distance is larger than the preset threshold, an interpolation point is inserted between the two adjacent touch points; and S3 drawing a preset writing brush shape according to points on the touch track to generate handwriting. The handwriting generating method enables writing brush handwriting input by hands to be smooth and continuous.

The embodiment of the invention provides a link failure processing method and a data transmitting device, wherein the method comprises detecting that a link goes wrong, starting a timer, sending out failure detecting conversation dropping notice, receiving a detecting message by the timer before time over, keeping the neighbor relationship with a neighbor route device connected through a link, generating route information, and sending out the route information. The invention simultaneously provides a data transmitting device for implementing the method. The embodiment of the invention can shorten the convergence time length of an upper-layer application module, improves the convergence property, and saves CPU resource of the data transmitting device, and improves the working property thereof.

Provided are a mobile terminal, device and method for controlling the waking of application programs in the mobile terminal, the method comprising the following steps: acquiring one or more application programs having a wake cycle and running in the background in the mobile terminal, and acquiring the identifier of each application program; according to the identifier of each application program, searching a preset wake cycle rule list to obtain the wake cycle of each application program; and waking the corresponding application program according to the wake cycle. The method of an embodiment of the present invention reduces wake times and energy consumption, and improves the life of the battery of a mobile terminal, while saving CPU resources and improving the performance of the mobile terminal.

The invention discloses an image processing method and device in GPU vitualization, and relates to the technical field of GPU vitualization. The image processing method and device in GPU vitualization solve the problem that because a front-frame image and a rear-frame image are compared through a CPU, burdens of the CPU are large in the prior art. The method comprises the steps that the current-frame image is obtained through a GPU; differentiated comparison instructions are received, and a front-frame image is obtained; according to the differentiated comparison instructions, a difference image between the current-frame image and the front-frame image is obtained and the position of the difference image is obtained; the position of the difference image and the difference image are sent to a system memory. The embodiment of the image processing method and device is applied to the image processing process in GPU vitualization.

The invention discloses a method for memorizing data and a network management system, wherein the method comprises the steps: a performance data base builds a performance table which corresponds to collected granularity according to the different collected granularities and builds an index; a performance server collects performance data from a network element, memorizes the performance data into corresponding positions of the performance table, and generates an index; and the performance server queries the performance table and transmits a query result to a client end when the client end requests to query the performance table or the performance data. By building the performance table of the different collected granularities in the performance data base and building the index, the method eliminates unnecessary fields and redundancy information, thereby saving memory space, reducing the use of high-speed memory space when memorizing mass performance data, improving the speed, and saving the memorizing resources and CPU resources.

The invention discloses an island detection method and system, relating to the technical field of inverter control. The method comprises the following steps of acquiring a voltage signal of a grid-connected point of a generating system, extracting the phase information of the voltage signal; establishing a sliding mode frequency offset island detection curve by adopting a quadratic function; and generating a disturbing signal according to the sliding mode frequency offset island detection curve and sending the disturbing signal to an inverter of the generating system. The method and system provided by the invention can be used for avoiding detection blind spot, and moreover, the control is simple, and the inverter island can be fast and accurately detected.

The invention discloses a video playing method and device, a terminal and a storage medium, and the method comprises the steps: responding to a received first-level page playing instruction for a target video stream in a video aggregation page, and obtaining the data information of the target video stream; creating a target player instance according to the data information, and storing the targetplayer instance into a preset player management component; establishing a first association relationship between the target player instance and a first playing window in the video aggregation page; inresponse to a received secondary page switching instruction for a target video stream being played in the video aggregation page, releasing the first association relationship; and when switching to the target secondary page, establishing a second association relationship between the target player instance and a second playing window in the target secondary page. According to the invention, the video aggregation page and the target secondary page share the target player instance, so that the effect of switching pages without stopping playing is realized, and the watching experience of a user is more continuous.

The invention provides a resource-saving cloud service system and a resource saving method. The cloud service system comprises a cloud server and a cloud client, wherein the cloud server comprises n service nodes, a node management module, a data management module and a node control module, and is used for mounting user data, encoding and sending video streaming, returning, receiving and storing the user data; the cloud client realizes connection and information transmission through the internet and a certain service node of the cloud server. According to the method, through a timeout sleep and timeout turnoff two-stage energy-saving mechanism, if a user leaves for a short time, cpu (central processing unit) consumption is reduced through sleep, and quick connection can be guaranteed after the user comes back; if the user leaves for a long time, consumption of a cpu and a memory is reduced through turnoff. A dynamic sharing mode is adopted among the service nodes for cpu resource consumption of the cloud server, and when one service node is reduced in cpu consumption, saved cpu resources can be used by other nodes immediately.

The embodiment of the invention provides a method and a system for capturing network data. The method comprises the following steps: capturing a data packet by a network card, and buffering into a receiving first-in and first-out queue of the network card; transmitting the data packet in the queue to a pre-configured circular buffer region through a PCI DMA (peripheral component interconnect direct memory access) controller; on the basis of a preset interrupt trigger strategy, transmitting an interrupt request to a kernel mode drive; stopping receiving a new interrupt request which is transmitted by the network card by the kernel mode drive according to the interrupt request, updating descriptors of the circular buffer region, and restoring to receive the new interrupt request transmitted by the network card; extracting the data packet when a user mode data analysis device judges that the data packet exists in the circular buffer region according to the current descriptor of the circular buffer region, and updating the descriptor of the circular buffer region, thus zero copy of the overall transmission process is achieved. A system call is not required, so that resources of a CPU (central processing unit) are saved; the transmission efficiency of the data packet is improved.

The invention discloses a substrate offset diagnosis and correction method, which comprises a diagnosis step (10) and a correction step (20) as follows: in the diagnosis step, determining whether the substrate carried by a transmission part is offset based on the detection result of a first sensor; in case of no offset, do not carry out the following correction step; in case of offset, carrying out the correction step; in the correction step, adjusting the location of the substrate, and then enabling the substrate to eliminate offset and return to the baseline position. In addition, the invention also provides a substrate offset diagnosis and correction device which comprises a diagnosis unit and a correction unit, wherein, the diagnosis unit is used to judge whether the substrate carried by the transmission part is offset; in case of offset, triggering the correction unit to work; in case of no offset, do not carry out any operation; the correction unit is used to adjust the location of the substrate and enable the substrate to eliminate offset and return to the baseline position. The method and the device are simple to use and enable the substrate to be accurately transmitted; besides, the device has high transmission accuracy and good reliability.

The packaging retransmission method of message in network communication includes the following steps: firstly, the tunnel interface information required for packaging GRE protocol message is broadcasted to interface plate by means of main control board in the network equipment, the tunnel interface information includes source address, destination address and tunnel identification, then the interface plate uses the above-mentioned tunnel interface information to package said GRE protocol message, and finally the packaged GRE protocol message can be retransmitted by means of interface plate.

The invention discloses a multicast replication method and device in an Overlay network. The method comprises: determining the types of nodes at first, wherein the types of the nodes comprise a root node and a non-root node, and the root node at least comprises one of the following: a subnet root node, a station root node and a global root node; sending an IP address to the nodes according to the types of nodes and an obtained network topology, and notifying the nodes to establish a direct tunnel according to the received IP address, for enabling the nodes to forward a multicast message through the direct tunnel after receiving the multicast message. The multicast replication method and device disclosed by the invention can be used for reducing the replication of multicast flow in the Overlay network, saving the CPU resources of head end equipment and saving the network bandwidth of a data center.

The route device includes route management unit and ISIS protocol unit. Virtual ID assignment table in use for coincidence relation between pointed route protocol and virtual ID assignment range is configured in ISIS protocol unit. When ISIS protocol unit generates LSP, the coincidence relation is in use for setting up restriction for taken value of virtual ID based on type of protocol of LSP. Moreover, when receiving LSP, the coincidence relation is standard of judgment based on virtual ID to determine protocol type of LSP. The invention also discloses smoothing restarting method, which expands current well-known smoothing restarting methods. The invention requires that after accomplishing GR and after determining that the introduced other route protocols also accomplish GR, the said ISIS protocol unit just starts to update LSP corresponding to other route protocols. The invention solves issue of topology concussion of network route caused by introduced routes in GR procedure effectively.

The invention relates to a rapid route convergence method. After any routing node in a network receives a new link status packet LSP, the routing node receiving the new link status packet LSP does not start a Dijkstra algorithm immediately to conduct route recalculation of a whole topology, and the routing node analyzes change situations of the network topology according to a current shortest path first tree SPT, a link status information base LSDB information and information of an adjacent node contained in the new link status packet LSP, and the change situations of the network topology are divided into the following four classes: first, one leaf node is added in the network; second, the network topology has no changes and any parts of routing information is changed; third, a link is broken; and fourth other situations exist. The change situations of the network topology determine nodes participating in storage protect feature (SPF) calculation and calculate the route. The rapid route convergence method optimizes calculation and updating of the algorithm, can rapidly responds to topology changes and can reduce central processing unit CPU load.

The invention provides a method and apparatus for suppressing neighbor oscillation. The method comprises the following steps: a first router obtains a neighbor state change dynamic state between a second router which establishes a neighbor relationship with the first router; the first router adds one to a neighbor penalty value when the neighbor state change dynamic state is changed from a normal state to a fault state each time; when the increased neighbor penalty value is greater than a preset suppression threshold value, the first router sets the neighbor relationship as a suppression state. The method and apparatus solve a problem that an unstable neighbor state causes routing oscillation of a local routing table in the prior art so as to achieve an effect of saving valuable bandwidth resources and CPU resources.

The invention provides a graph drawing method and system based on the WPF technology. The graph drawing method includes the steps that firstly, a background thread receives time domain data collected by a data collection device; secondly, the time domain data are processed into frequency spectrum data by the background thread through fast Fourier transformation; thirdly, the spectrum data are mapped by the background thread through a set color gradation so as to obtain the RGB value, corresponding to a complete frequency spectrum line, on a displayer; fourthly, the frequency spectrum line is stored into a cache; fifthly, whether the cache is full or not is judged, the sixth step is executed if the cache is full, and otherwise the third step is executed; sixthly, a foreground main thread is informed by the background thread to draw the frequency spectrum line. The graph drawing method and system based on the WPF technology have the advantages that when a Doppler frequency spectrum graph is drawn, the consumption of CPU resources can be greatly reduced, and working efficiency can be improved.

The invention discloses a video transcoding method and device and a system. A graphics processor GPU is used to decode an input code stream to acquire decoded input code stream. A central processing unit CPU is used to analyze the decoded input code stream to acquire macroblock information needed by coding; and according to the macroblock information needed by coding, the central processing unit CPU is used to code the decoded input code stream according to a target code rate. GPU decoding is used to save CPU resources. Code stream analyzing is carried out on the decoded stream. Without affecting the transcoding quality, the overall transcoding speed is improved. Transcoding CPU resources are saved.

The invention discloses a method and device for synchronizing security association (SA) between equipment. The method comprises the following steps of: receiving identification information of opposite-terminal IPsec (internet protocol security) SA peer equipment transmitted from the opposite-terminal IPsec SA peer equipment in an IKE (Internet Key Exchange) SA negotiation process between IPsec SA peer equipment and the opposite-terminal IPsec SA peer equipment; deleting the IKE SA and the IPsec SA, inquired according to the identification information, on the IPsec SA equipment; and establishing the IPsec SA between the IPsec SA peer equipment and the opposite-terminal IPsec SA peer equipment by virtue of taking the identification information as an index. The method and the device can be used for solving the problem of data loss between the equipment, saving CPU (Central Processing Unit) resources and improving the utilization rate of the CPU resources; and as the IPsec SA peer equipment is informed of deleting the useless SA by an IKE SA negotiation message in the IKE SA negotiation process, the burden from extra message interaction between the equipment is not added, and bandwidth resources are saved.

The invention discloses a method of simultaneously realizing the functions of two-layer exchange, three-layer routing, data content filtering and firewall in a chip, using a single chip to realize multiple network functions, and its whole technical solution: 1. making network address transform (NAT); 2. making two-layer exchange; 3 making three-layer routing. It can reduce networking expenses, saves CPU resources, makes CPU specialize in other functions, and improve the whole property.

The embodiment of the invention discloses a method, device and system for loop path detection. The method for loop path detection comprises the following steps that a loop path existing in a network element is obtained from topological data of a port of the network element, the loop path is stored in a loop path table, and the topological data comprise identification information of the port and identification information of a next-level node connected with the port; when a medium access control address of the port of the network element is shifted, the loop path table is looked up according to the original identification information of the port and the shifted identification information, and the loop path on a protocol is obtained.

The invention provides a data message transmission method, a data message transmission device and a machine readable storage medium. The method comprises the steps of after a data message is received,querying a first FIB table via a target address of the data message; if the FIB table does not have a forwarding table item corresponding to the target address, sending a request message carrying thetarget address to computing equipment, wherein the request message is used for allowing the computing equipment to query the forwarding table item corresponding to the target address from the secondFIB table of the computing equipment, and returning the queried forwarding table item; and receiving forwarding table item corresponding to the target address sent by the computing equipment, storingthe received forwarding table item into the first FIB table, and using the received forwarding table item to send the data message. According to the technical scheme provided by the invention, the processing pressure of the network equipment can be relieved, the CPU resources of the network equipment can be saved, the problem of high CPU utilization rate is avoided, and the stability of the network equipment is improved. The utilization rate of the table item resources is improved, and the service experience of the user is improved.

The invention provides a method for optimizing network throughput in a virtualization environment of an embedded network. According to the method, a plurality of hypercalls are aggregated into one hypercall through an aggregation timer, the aggregation interval time can be automatically adjusted according to the arriving time of a predicted next I / O request, in this way, the frequency for context switching of a client and a host machine is remarkably lowered, and a lot of CPU resources are saved so that more network I / Os can be processed; meanwhile, a delay of data transmission can be reduced under the condition of low network traffic. Through experiments of the inventor, it can be found that compared with a traditional method, the AHC method enables the throughput of Netperf, the throughput of Apache and the throughput of Memcached to be increased by 221.14%, 146.23% and 257.42% respectively.

The invention discloses an information interactive system and method based on PCI (Peripheral Component Interconnect) and an upper computer. The system comprises an upper computer and PCI motion control equipment, wherein the PCI motion control equipment comprises an interruption triggering unit used for trigging PCI interruption while the PCI motion control equipment has a message to be reportedto the upper computer, wherein an interruption event of the upper computer preset in a PCI driving program is triggered by the PCI interruption; and the upper computer comprises an event sensing unitused for sensing the interruption event by using a preset thread and an event processing unit used for correspondingly processing the event after the interruption event is sensed by the event sensingunit. By using an embodiment of the invention, the upper computer can obtain the message of the PCI motion control equipment in time and correspondingly process the message; moreover, the thread established by the upper computer substantially does not occupy the space of a CPU (Central Processing Unit) so that the CPU resources of the upper computer are largely saved.

The invention discloses a method for generating a photographing associated short video through a picture pick-up device and downloading and playing the short video. The method comprises the following steps of (1) generating a short video record, wherein the picture pick-up device simultaneously takes a picture and generates a data record, and a file name of the short video, a file name of an associated long video, and a starting time and an end time of the short video are recorded in the data record, so that the associated short video is stored in the picture pick-up device in a form of one data record; and (2) downloading the short video, wherein a video playing client applies to the picture pick-up device for downloading the short video, the picture pick-up device searches a corresponding file record, intercepts the short video from the corresponding long video according to the starting time and the end time of the short video, and then sends to the video playing client to play. The method has the advantages that duplicate record of the same content can be avoided to save the storage space of the picture pick-up device, and the resources of a CPU (Central Processing Unit) can be saved during picture recording.

The invention discloses a graphics interchange format (GIF) file processing method and a GIF file processing system for a digital television system. The processing method comprises a front-end processing method and a terminal processing method. The front-end processing method comprises the following steps of: decoding a GIF file to obtain a red, green and blue (RGB) image frame and the transparent information of pixels of the RGB image frame; regulating RGB values of the pixels according to the transparent information of the pixels of the RGB image frame, wherein the regulated RGB values are used for reflecting whether corresponding pixels require transparent processing or not; converting the RGB image frame with the regulated RGB values into a luma and chroma (YUV) image frame; performing video coding on the YUV image frame according to a preset video coding standard to obtain video data; and transmitting the video data. The device provided by the invention comprises a front-end processing system and a terminal processing device. By the processing method and the processing device provided by the invention, the finite central processing unit (CPU) resources of a digital televisionset-top box are saved, the GIF file with a high frame rate and large images can be used, and dynamic images, with transparency effects, of the GIF file can be displayed at a terminal.

The invention provides a dual-controller communication method, a transmitting end controller and a receiving end controller. The dual-controller communication method includes the steps that according to the position and the size of a data written-in magnetic disk for data needing to be transmitted, the transmitting end controller applies a receiving buffer pool of the receiving end controller for receiving buffer space and obtains a buffer position pointer of the receiving buffer space; the transmitting end controller transmits the data needing to be transmitted to the applied receiving buffer space of the receiving end controller, and transmits the buffer position pointer of the receiving buffer space to the receiving end controller. The receiving end controller stores the received data in the receiving buffer space, and converts the received buffer position pointer into a local buffer data pointer, and transmits the data pointer to an RAID system stage by stage from a rolled data-caching mechanism. By means of the technical scheme, CPU resources of a computer and a bus bandwidth of an occupation system are saved, and the system processing performance is accordingly improved. The dual-controller communication method, the transmitting end controller and the receiving end controller are mainly suitable for dual-controller communication.

The invention discloses a method and equipment for using route switching equipment, and the method comprises the steps: responding to the switching of the route switching equipment from a conventional mode to an open flow mode, and carrying out the following steps: converting items of a forwarding table of the conventional mode into flow table items of the open flow mode; storing the converted flow table items to be stored in the flow table of the route switching equipment; responding to a received data message, and carrying out the processing of the data message according to the converted flow table items stored in the flow table. The method and equipment can enable the route switching equipment to be switched to the open flow mode from the conventional mode, and save the CPU resources of the route switching equipment.

The invention discloses a VNC-based multi-screen interaction method and device. The method includes: detecting screen change rate of a VNC server desktop; if the screen change rate detected is larger than or equal to a first preset value, detecting window change rates of windows in the VNC server desktop one by one; if the window change rates of the windows are larger than or equal to a second preset value; encoding the windows by H264, JPEG or MJPEG; transmitting the windows encoded to a VNC viewer desktop. The second preset value is larger than the first preset value. The invention further discloses the VNC-based multi-screen interaction device. Through the application of the VNC-based multi-screen interaction method and device, the problem that the VNC server desktop changes fast, for example, images obviously jam or slow down when video or animation is played, is solved effectively.

The invention discloses a kitchen fire alarm system and method for human body detection in front of a cooking bench. The system comprises a flame detection module, an analog-to-digital conversion module, a data transmission module, a main control module, a combustible gas detection module, a GSM communication module, a human body detection module, a temperature detection module and a distance measurement module. The method comprises the steps that the main control module calculates a combustible gas concentration value, a flame concentration value and a temperature value and compares the values with corresponding alarm thresholds; when any numerical value of the combustible gas concentration, the temperature and the flame concentration is detected to exceed the preset alarm thresholds, thehuman body detection module detects infrared rays emitted by a human body in front of a cooking bench, the distance measurement module measures the distance between the head of the human body in front of the cooking bench and the alarm system, and the main control module calculates the height of the human body in front of the cooking bench and selects whether to send alarm information or not. Before the kitchen fire alarm information is sent, the human body in front of the cooking bench is detected, and the practicability of the kitchen fire alarm system is improved.