60results about How to "Solve Bandwidth Bottlenecks" patented technology

The invention discloses a big data storage and access system and method based on a distributed file system. The big data storage and access system based on the distributed file system comprises the distributed file system which comprises a metadata server, a storage server cluster, an access client and a log server, the metadata server manages and monitors metadata information in the distributed file system and the storage server cluster, the storage server cluster stores data contained in the metadata information in multiple data storage nodes of the storage server cluster, the log server backs up the metadata information in the metadata server and synchronizes operation logs of the metadata server in real time, and the access client is used for selecting standby metadata service processes in the metadata server or in the log server according to the connection state of the metadata server to store the server cluster.

A pixel pre-fetching device of movement compensation module in AVS video hardware decoder consists of decision control module, write buffer storage unit, read buffer storage unit, gating unit, forward pre-fetching buffer storage, backward pre-fetching buffer storage and reread buffer storage. The method utilizing said device to pre-fetch pixel is also disclosed.

The invention relates to a multi-link concurrent access multimedia data transmission method. The present method that using a single operator to provide transmission bandwidth can not support the high quality and real time transmission of multimedia service. The invention provides a multi-link concurrent access multimedia data transmission method which includes the server information acquisition, the segmentation of the data to be transmitted, the data transmission, and priority distinguished data transmission and data protection. Wherein, the data transmission includes an alternative test cycle and a data cycle; in the test cycle, the transmitting equipment test the performance of the data transmission link and the access link; in the data cycle, the transmitting equipment transmits data at each link according to the test result of the test cycle, thereby making full use of the link bandwidth resource and avoiding the link congestion. The multi-link concurrent access multimedia data transmission method of the invention adopts the transmission type by distinguishing the priority of different data at different priority; thereby ensure the transmission reliability of the high priority data.

An image element cache method and system, which is used in the motion compensation of an audio decoder in the field of audio decoding, includes: step 1. Reading motion compensation parameter information from the parameter memorizer; step 2. Zoning the area of the reference frame where the current basic block of motion compensation needs; step 3. Reading and selecting the image elements that are not in the internal memorizer out of the external memorizer and memorizing them in the internal memorizer; step 4. Reading and selecting out the data in the reference frame image element unit of the reference frame area from the internal memorizer in turn, and sending them together with the current motion vector of the basic block to an external interpolator. The invention system includes: an overall controlling module, a parameter module reading motion compensation, a reference frame area module which the motion compensation need, a judging and replacing module of image element unit, a controlling module of interpolation data output, and a memorizing module of data and label. The invention reduces the bandwidth need for external memorizer, and resolves the bandwidth bottleneck problem of the external memorizer of the audio decoder design.

The present invention discloses an access method of Internet internal network Web service comprising the steps of setting one or more hardware servers in the Internet as channel server, dynamic domain name server and management server; the management server receives the request of establishing channel of client for providing Web service through the Internet, and judges whether the client has external network IP address; if the client has external network IP address, starts the channel server, and continues receiving the requests of other clients; if the client is internal network computer, the client chooses one channel server and establishes data channel with the channel server; synchronously , the dynamic domain name server updates the IP address information. The internal network Web service access method of the invention is able to effectively penetrate each Middle-boxes limitation in Internet, and break port locking.

The invention provides a multi-die higher-order photonic switching structure based on a high-density memory in order to solve the problem of limited structure of a switching chip due to the bandwidthbottleneck of the memory of the conventional photonic switching structure. The structure is formed by mutual connection of K completely same photonic switching dies based on the high-density memoriesthrough an inter-chip interconnection network and includes N input / output ports, and each photonic switching die comprises N / K input ports, N / K electro-optical conversion modules, N / K separators, a storage input photonic switching network, an optical waveguide, a storage input buffer, a memory control logic, an inter-chip interconnection input output agent module, an inter-chip input buffer, an alternative selector, the high-density memory, a storage output buffer, a storage output photonic switching network, N / K photoelectric conversion modules, and N / K output ports, wherein K, N and k are all integers, 1<=k<=K, and K<=N. According to the structure, the complexity of the switching structure and the cost of optical devices are substantially reduced, and the order of the switching structureis further extended.

The invention relates to a multi-version GIS slicing service system which comprises a file storage server, an application cache server, a slicing server and a client, wherein the file storage server is used for storing slicing associated files in layers according to types and versions, wherein the slicing associated files include a slicing data file, a slicing resource description file and a slicing index file, the slicing data file is used for storing slicing data information, the slicing resource description file is used for describing the type, grade, coordinate and version information of the slicing data, and the slicing index file is used for describing the index information of the slicing data; the application cache server is used for loading the slicing index information into the application cache server; the slicing server is provided with an access interface capable of receiving a slicing request and is used for receiving a slicing request of the client, acquiring the corresponding slicing index information and slicing data and returning the corresponding slicing index information and slicing data to the client; the client is used for sending the slicing request and receiving the returned slicing data. The system is capable of providing access and management for multiple versions of GIS slicing, and is simple in system structure and good in use effect.

The invention relates to a statistical function distributed execution method for edge computing, which comprises the following steps of: registering Internet of Things equipment as a leaf node of an Internet of Things system through an edge gateway of the Internet of Things system; caching data resources collected by the leaf node to the edge gateway; sending a query request for the data resourcestatistical value sent by a user to a cloud server of the Internet of Things system in a URL expression form; enabling the cloud server to carry out address mapping on the query request and forward the query request to a corresponding edge gateway; selecting corresponding data resources from the cache data stored in the edge gateway and returning the selected data resources to the cloud server; and arranging the intermediate result at the cloud server to obtain a statistical result and feeding the statistical result back to the user.

The invention provides a memory controller based on the burst disorder memory access dispatching, which is used to the burst disorder memory access dispatching accessed by a memory, and comprises a reading-writing queue module for storing the reading-writing memory access from a processor in a two-dimension mode, a block arbitration module for arbitrating a burst memory access from each block in each memory clock period, and an event selection module for selecting the final memory access operation from the burst of each block after arbitrated to send to the memory. The block is conducted with the burst arbitration to dispatch the disorder memory access through changing the structure of a queue and adopting the priority expression provided by the invention, thereby increasing the data bandwidth of the memory, and reducing the executing time of the processor.

The invention discloses a media distribution system and method for an intelligent terminal, and relates to the field of media distribution. The system comprises the intelligent terminal, a management server and a content server, wherein the intelligent terminal comprises a platform interaction module, a download management module and a media play module; the management server comprises an authentication distribution module, a user database, a media configuration database and a log database; each message sent to the management server by the intelligent terminal at least carries a hardware unique identifier corresponding to an intelligent terminal device, a user name or a password; and the content server stores to-be-downloaded resources. According to the media distribution system and method, the manpower cost of maintenance can be reduced, the demand on the network bandwidth and the requirement on the server performance can be lowered, the bandwidth bottleneck caused by concurrent access can be avoided, the media content can be obtained in time, and the playing quality is relatively high; and the advertisement deployment cost can be reduced, so that the media distribution system and method can be widely applied to places such as shopping malls, cinemas, guesthouses, hotels, office buildings and the like.

The invention provides one network GIS system and its data transmission method, wherein, the system comprises the following parts: GIS customer end to provide GIS service require operation and to get service result; application servo group to respond GIS service require and services; subject based memory group to store subject space data and to provide common pool to application servo. The transmission method comprises the following steps: space data stores space memory subject based on group and the application servo group and the memory group are combined.

The invention provides a multi-frequency RFID reader based on software radio. Due to lack of a uniform standard, the radio frequency recognition frequencies of all countries are not consistent, and in order to adapt to the requirements of different countries, a reader-writer is required to have an RFID product which has the requirement of recognizing multiple frequencies to serve for different countries or districts. The invention provides the multi-frequency RFID reader based on software radio. The multi-frequency RFID reader is formed by a radio frequency module, a baseband module and a power module, wherein the power module is respectively connected with the radio frequency module and the baseband module; the radio frequency module is used for realizing modulation and demodulation on a baseband signal and transmitting or receiving a modulated signal; the baseband module realizes control, encoding, decoding and inspection of the whole reader-writer, interaction of the reader-writer and an upper computer, and control on the radio frequency module; and the baseband module comprises an ARM processor system and a programmable logic, the ARM processor system is connected with the programmable logic and the programmable logic is connected with the radio frequency module.

The invention provides a steam turbine shafting vibration data acquisition, preprocessing and remote sending system. The system includes a key phase signal sensor, a key phase signal conditioning circuit, at least one vibration signal sensor set, at least one vibration signal conditioning circuit, a pulse counter, a multichannel synchronous digital-to-analog converter, an ARM controller, an SRAM,an FPGA high-performance calculation module, an Ethernet communication module and a cloud server. According to the system provided by the invention, multichannel synchronous steam turbine shafting vibration data acquisition and complicated real-time preprocessing and sending functions are integrated in one system, aiming at steam turbine shafting vibration acquisition, preprocessing and remote sending service requirements based on internet and big data, based on actual need of data analysis, real-time preprocessing and feature extraction are performed after original vibration data are acquired, and feature values required by data analysis are remotely sent by use of a network.

The invention discloses an efficient and safe multi-keyword-supporting searching method in a mobile cloud. The method is based on a TF-IDF (Term Frequency-Inverse Document Frequency) search model; a data owner stores an encrypted document and an encrypted document index in a cloud end, and a data user inputs a keyword which is encrypted by Fast Accumulated Hashing to search the encrypted document stored in the cloud end, so that search safety is realized. The isolation of search is realized through the addition of noise. The search control is realized through user registration and verification. The high efficiency of search is realized through one-time communication, keyword compression, establishment of search cache and binary set. The efficient and safe multi-keyword-supporting searching method in the mobile cloud guarantees security of document collection, keywords and indexes involved in the search process in the aspect of security, simplifies the search process, increases the search efficiency, compresses total communication amount and reduces communication bandwidth in the aspect of efficiency, and solves the bandwidth bottleneck problem of the encrypted search on a cloud platform.

The method includes following steps: the entitlement control message (ECM) transmitted by broadcast front-end equipment includes composite control message about conditional access; the composite control message is composed of non-randomness conditional message and randomness message to construct the data relating to cryptographic algorithm of scramble / descramble control word; the conditional message is a conditional reception control message relating to active reception time field or time quantum; the smart card in the reception end uses cipher key and conditional message in composite control message to compare and renew the relevant control message in smart card, and decides if the entitlement control message or composite control message are processed by using a certain kind cipher algorithm; the correct scramble / descramble control word is obtained to make descrambling for received scramble / descramble transport stream (TS).

The embodiment of the invention provides a data processing method, device and system, and the method is used for an upper server in a data processing system. The data processing system also comprisesat least one lower server for storing the data contents of a corresponding region. The method comprises the steps: receiving a data read / write request transmitted by a target terminal; obtaining the target user identification information carried by the request; determining a target communication identification corresponding to the target user identification information according to the preset mapping relation relative to the user identification information and the communication identifications of all lower servers; enabling the request to be redirected to the target lower server correspondingto the target communication identification, so as to enable the target terminal to read / write the data in the target lower server according to the data read / write request. According to the embodimentof the invention, the method reduces the storage pressure of the upper server and the data access pressure, and prevents the network bandwidth neck of the upper server.

The invention relates to a transmission method of broadcast quality audio and video based on a commercial wireless network. At present, the wireless live rebroadcast technology has great limitations.The method of the invention comprises the following steps: firstly receiving a mixed audio and video steaming in a camera; transmitting the mixed audio and video streaming to a standard encoder; carrying out compression coding on the mixed audio and video streaming in the standard encoder to obtain a compressed TS code stream; processing the compressed TS code stream to obtain a multi-channel codestream; transmitting the multi-channel code stream into an objective server through the commercial wireless network and Internet at real time in a concurrent manner; rearranging and recovering the received multi-channel code stream into a complete compressed code stream by the objective server, and transmitting the complete compressed code stream to the standard encoder; and carrying out decodingand uncompressing processes on the recovered compressed code stream through the standard encoder to obtain the audio data stream and video data stream shot by a broadcast quality camera. The method of the invention can simultaneously switch in multiple commercial wireless networks and Internet, transmits the audio data stream and the video data stream shot by the broadcast quality camera into television stations at real time, thus improving the instantaneity of news.

The embodiment of the invention discloses a data processing method and a server, and aims to solve the problem of network bottleneck due to centralized issuing of container mirrors and optimize the container service quality. The method in the embodiment of the invention comprises the following steps: acquiring a target distributed mirror warehouse node; determining a target centralized warehouse node according to the target distributed mirror warehouse node, wherein the target centralized warehouse node is used for storing all container mirrors in a target container mirror warehouse; storing target container mirrors in the target distributed mirror warehouse node in a distributed way, wherein the target container mirrors are container mirrors of which usage rates reach a preset value in the target container mirror warehouse; receiving a target request, wherein the target request is used for requesting issuing of a first container mirror; judging whether the usage rate of the first container mirror reaches the preset value or not; and if so, issuing the first container mirror from the target distributed mirror warehouse node.

The embodiment of the invention discloses a data packet forwarding method, device and system. The method comprises the following steps that the corresponding relationship between a user and a forwarding mode sent by an AC (access controller) is received; the corresponding relationship between the user and the forwarding mode is determined by the AC according to the forwarding mode corresponding to a user group; the forwarding mode comprises a local forwarding mode and an integrated forwarding mode; an uplink data packet is received, wherein the uplink data packet carries a user identifier; according to the user identifier, the uplink data packet is determined to be from a first user; the forwarding mode corresponding to the first user is determined according to the corresponding relationship between the user and the forwarding mode; the uplink data packet is forwarded according to the forwarding mode corresponding to the first user. The method, the device and the system provided by the embodiment of the invention have the advantages that the uplink data packet of the user is forwarded according to the forwarding mode corresponding to the user on an AP (access point), and the bandwidth bottleneck problem of the AC in the integrated forwarding mode can be avoided.

The invention discloses a high-definition decoder digital display mixed format bitstream self-adapting processing system and method. The high-definition decoder digital display mixed format bitstream self-adapting processing system comprises a decoder, a DDR2 controller, a motion vector up conversion module, a display control module and an output playback module; wherein the decoder is used for reconstructing and decoding the mixed format image to obtain the mixed bitstream; the DDR2 controller is used for the protocol control of the DDR2 physical layer protocol PHY circuit, and used for pre-storing the decoded mixed bitstream to the DDR2 / SDRAM off chip physical memory to realize the data read and write to the DDR2 / SDRAM off chip physical memory; the motion vector up conversion module comprises a direct memory access unit, a high definition motion vector up conversion queue buffer unit and a high definition motion vector up conversion unit; the display control module performs the cyclic redundancy check of chrominance / luminance pixel components; and the output playback module is used for outputting the image to the off chip interface for display. The invention has the advantages such as the mechanism of the invention does not take advantage of the extra time of the CPU, the bus access width of the mixed pixels is fixed, and the format and order of the access pixels can be configured before de-interleaver.

The invention relates to a load balancing method and a system. The load balancing method comprises the steps: a router selects a first load balancer from a first load balancer group; the first load balancer receives a first data packet sent by the router and performs packaging treatment to obtain a second data packet; a load balancing manager distributes a virtual network address for a second load balancer in a second load balancer group, inquiries a first IP address and first port information of the second load balancer and sends the first IP address and first port information to the first load balancer; the first load balancer sends the second data packet to the second load balancer according to the first IP address and first port information; the second load balancer processes the second data packet to obtain a third data packet. The system comprises the router, the load balancing manager, the first load balancer and the second load balancer. The bandwidth bottleneck problem of the load balancer groups under a large-scale scene can be solved.

The invention is applicable to video monitoring, and provides a write-in method and apparatus for monitoring data. The write-in method for monitoring data comprises: obtaining a current write request; detecting whether files in the current write request and in a previous write request belong to the same group; if so, detecting whether a remaining space is left in a data region of a current storage block; and if so, writing monitoring data corresponding to the current write request into the data region of the storage block according to the sequential write mode. The invention changes the write operation mode of a video monitoring device for a storage medium from random write into sequential write, effectively avoids the bandwidth bottlenecks of the random write mode, and can also meet the ever-increasing write-in bandwidth requirement of video monitoring applications without additional hardware investment.

The invention discloses a system and a method for realizing multi-picture based on a serial differential switch. An image processing unit for implementing the method comprises a plurality of decoding modules and encoding modules which are connected with each other through serial differential interfaces. The method comprises the following steps: when each of the decoding modules is connected with the serial differential switch through the serial differential interfaces, zooming received image data to small picture image data in multi-picture image data by the decoding modules; transmitting the small picture image data to the encoding modules through the serial differential switch; and combining a plurality of small picture image data into multi-picture image data by the encoding modules. Because the transmission of image data is realized by the serial differential switch, a bottleneck problem of video bandwidth is solved so that a conference television user can receive high-definition multi-picture clearly; and simultaneously, the number of digital signal processors can be selected as required according to the maximum number of the multi-picture, which not only saves implementation cost for low-end users, but also meets the requirements of high-end users by increasing the number of the digital signal processors.

The invention discloses a method for applying a high-performance server to a vehicle-mounted self-organizing network, and belongs to the field of server application and Internet of Vehicles. The technical problem to be solved by the invention is how to use a server as an RSU infrastructure. The flexibility of the RSU node in the VANET is improved, and cost is reduced. Therefore, more RSUs are arranged to ensure the connectivity of the data and the selectivity of optimal calculation. The adopted technical scheme is as follows: according to the method, a high-performance server is used as a roadside unit, and a macroscopic path selection mechanism based on roadside unit assistance is adopted. The distribution condition of vehicles on a road is obtained through information interaction betweenvehicle nodes and roadside units deployed on the roadside and multi-class V2R message statistics. Therefore, the road density is calculated, and an optimal transmission path from the source node to the target node is selected by adopting a global design idea based on the principle of ensuring connectivity and short path priority.

The invention discloses a data processing method and device, a processor can comprise a cache module and a plurality of logic modules, the processor can obtain image data collected by an image sensorand store the image data in the cache module, and when the image data is processed through a data processing model, for each layer of the data processing model, the data processing module processes the image data. Model parameters of the layer are acquired, the model parameters are loaded to each logic module, an intermediate image obtained from the upper layer of the layer is acquired from the cache module, the intermediate image is segmented to obtain each sub intermediate image, each sub intermediate image is loaded to each logic module, operation on the model parameters and each sub intermediate image are performed through each logic module to obtain a plurality of sub intermediate images, an intermediate image obtained by the layer is acquired, and the intermediate image is stored ina cache module. By means of the method, the image data or the intermediate images obtained in all the layers are stored in the cache, the problem that the data transmission bandwidth is limited is solved, computing power resources of the processor are effectively used, and the effect of reducing time consumption is achieved.

The invention discloses a multi-hop cluster type wireless sensor network system based on millimeter waves. The multi-hop cluster type wireless sensor network system comprises a cluster head node, data collecting nodes, a millimeter wave routing node and a data center. The cluster head node and the multiple data collecting nodes form a cluster, the data collecting nodes collect data and transmit the data to the cluster head node, and the cluster head node collects the data through the cluster, transmits the data to the millimeter wave routing node and then converges the data to the data center. The multi-hop cluster type wireless communication system based on the 60 GHz millimeter waves is designed, a traditional wireless sensor network is adopted in intra-cluster communication, and 60 GHz millimeter wave multi-lop transmission is adopted in inter-cluster communication. The bandwidth bottleneck problem of the large-scale inter-cluster communication of the wireless sensor network is solved fundamentally.

The invention discloses an acceleration device, and relates to the field of computers, the acceleration device is applied to a CPU, the acceleration device comprises a GPU controller, a GPU and an IB exchange interface, wherein the GPU controller is connected with the GPU through an NVLink bus, the GPU controller is connected with the IB exchange interface through an IB bus, the CPU can transmit a data task needing to be processed to the GPU controller through the IB exchange interface via the IB bus, the GPU controller transmits the data task to the GPU through the NVLink bus for data processing. The GPU controller can achieve conversion between an IB protocol and an NVLink protocol, the interconnection architecture effectively solves the problem of data transmission bandwidth bottleneck of 64GB / s of bidirectional PCIE bus bandwidth in an existing CPU+GPU architecture, and data is transmitted and exchanged between a CPU memory and a GPU memory to achieve bandwidth balance.

The invention discloses a resource scheduling method based on OpenFlow, a scheduler and a system thereof. The scheduler comprises the following steps: when a Reduce task to be scheduled arrives, current topology information of the cluster is obtained Determining an execution node for executing a Reduce task to be scheduled and a first transmission path for transmitting an intermediate result corresponding to the Reduce task to be scheduled to the execution node from an idle Reduce node according to the current topology information of the cluster; Scheduling a Reduce task to be scheduled to anexecution node, reserving bandwidth for the Reduce task to be scheduled according to the first transmission path and updating cluster topology information; The intermediate result corresponding to theReduce task to be scheduled is transmitted to the execution node according to the first transmission path, so that the Reduce task to be scheduled is executed by the execution node. The invention canminimize the sum of the use costs of the computing resources and the network resources in the Reduce stage, thereby improving the execution efficiency of the task.

The invention provides a memory controller based on the burst disorder memory access dispatching, which is used to the burst disorder memory access dispatching accessed by a memory, and comprises a reading-writing queue module for storing the reading-writing memory access from a processor in a two-dimension mode, a block arbitration module for arbitrating a burst memory access from each block in each memory clock period, and an event selection module for selecting the final memory access operation from the burst of each block after arbitrated to send to the memory. The block is conducted withthe burst arbitration to dispatch the disorder memory access through changing the structure of a queue and adopting the priority expression provided by the invention, thereby increasing the data bandwidth of the memory, and reducing the executing time of the processor.