98 results about "Bandwidth throttling" patented technology

The invention discloses an inter-satellite link load balancing method, system and device and a storage medium, relates to the field of satellite communication, and is used for solving the problem of network congestion of a core network user plane. In the application, an SMF network element receives inter-satellite link state information from each UPF network element; calculating a routing path from the source satellite node to the target satellite node according to the inter-satellite link state information under the limitation of inter-satellite link time delay and bandwidth; and distributing the routing path decision to the UPF network element of the corresponding satellite node. According to the invention, the load balancing of inter-satellite links can be realized, the user plane network congestion of the satellite-borne core network is avoided, and the service quality of the satellite-borne core network is improved.

One embodiment provides a graphics processor comprising a memory interface, a graphics core cluster including a plurality of graphics cores, and an interconnect fabric to interconnect a plurality of hardware clients including the plurality of graphics cores. The interconnect fabric include a plurality of fabric ports coupled with the plurality of graphics cores. A fabric port is configured to limit bandwidth available to an associated graphics core via a bandwidth throttler circuit coupled with the fabric port.

The invention discloses a federated learning image classification method based on time extension gradient cumulative average. Each client uses a local data set to train a local model to obtain gradient information of a current communication round, and uploads the gradient information to the central server after compression; the client continues to perform local training in a time period from the time when the client starts to upload the compression gradient information of the current communication round to the time when the server completes the compression gradient information aggregation of the current communication round; the central server side aggregates the compression gradient information to obtain average compression gradient information of the current communication round, and issues the average compression gradient information to the client side to update a local model of the (t + beta) th communication round; and repeating the steps until the training is completed, and inputting the to-be-detected image into the local model to obtain an image classification result. The method shows excellent performance under various data sets and different compression levels, has higher convergence speed and model precision, and solves the problem of low communication efficiency caused by network delay and bandwidth limitation in image classification of federated learning.

The invention relates to an optical modem connection scheduling method based on scene awareness and dynamic resource slicing. The method comprises the following steps: collecting application layer data in real time, analyzing application layer protocol characteristics, identifying an application scene, and predicting a bandwidth demand and a change trend based on the application scene; creating a private network slice for the emergency task; detecting a network flow mode in real time, acquiring an application type of the network flow mode, identifying a high-bandwidth downloading behavior, detecting a network index of a delay sensitive application in real time to determine a service level protocol, and determining a real-time application priority and a current demand according to the application type and the service level protocol; the required minimum bandwidth is calculated, dynamic speed limiting, bandwidth redistribution and network utilization rate monitoring are carried out on high-bandwidth downloading application, and downloading bandwidth limitation is temporarily improved when the total utilization rate is lower than a threshold value and continues for specified time; calculating a risk score; a heartbeat packet acceleration channel is set, an independent UDP packet forwarding path is developed for key IoT equipment, the connection state of the IoT equipment is detected in real time, and an emergency communication plan is triggered when the IoT equipment is abnormal.

The invention provides a soft-RoCE-based virtual network topology simulation method, which belongs to the technical field of network topology simulation, and comprises the following steps: S1, creating isolated virtual nodes; s2, a virtual network link attribute configuration step: link parameters are dynamically configured on a veth device by using a tcnetem tool, the parameters comprise time delay, packet loss rate and bandwidth limitation, parameter configuration supports real-time dynamic adjustment so as to meet simulation requirements of different topology scenes, and aiming at the problems that a traditional method is low in topology flexibility and has defects in multi-tenant isolation, the virtual network link attribute configuration method is provided for solving the problems that the traditional method is low in topology flexibility and has defects in multi-tenant isolation. According to the method, the GID, the IP and the protocol stack are isolated by flexibly configuring the link attributes and based on the network namespace, and the effect which cannot be achieved by a traditional method is achieved.

The invention discloses a model prediction direct position control method based on point-to-point trajectory planning, and belongs to the technical field of motor control. The method comprises the following steps: measuring a three-phase current signal, a rotating speed signal and a position signal of the PMSM; converting the three-phase current signal into a current component under a two-phase static coordinate system through Clark conversion; performing Park transformation to obtain a d-q coordinate system current signal; and performing voltage prediction to generate a d-q axis voltage control signal. Planning a target curve trajectory; and performing online prediction and rolling optimization. According to the invention, bandwidth limitation of a traditional cascade structure is broken through, position loop time delay is reduced, speed oscillation and position overshoot near a target positioning point are suppressed, stability of a control system is improved, multi-target collaborative optimization of position, rotating speed and current is realized, and economical efficiency and engineering practicability are both realized.

The invention particularly relates to a digital twin interactive media content streaming method in a weak network environment. According to the digital twin interactive media content streaming method in the weak network environment, a client sends a dynamic address request and receives fed back dynamic address information; the media server receives the SDP proposal and the ICE candidate sent by the client, and forwards the SDP proposal and the ICE candidate to the digital twin program; the client and the digital twin program establish a media content transmission channel with the media server according to the dynamic address information and the fixed address information; and the digital twinning program transmits the media content data to a client for display through a media content transmission channel stream by adopting a quality of service (QoS) processing mode. According to the digital twin interactive media content streaming method in the weak network environment, the problems of low quality and high time delay caused by bandwidth limitation and a severe network environment are solved, the quality of the media content received by the client is improved, the time delay in the transmission process of the media content data is reduced, and the transmission efficiency is improved.

The invention provides a data transmission method, device and equipment and a readable storage medium, and the method comprises the steps: monitoring the data flow of each calculation unit communicating with the outside, and configuring a data channel between a busy calculation unit and an idle calculation unit to be opened; issuing a first shunting instruction to the busy computing unit, and shunting the data message of the busy computing unit to the idle computing unit; and issuing a second shunting instruction to the idle computing unit, and forwarding the second shunting instruction outwards through a data channel of communication between the idle computing unit and the outside. According to the technical scheme of the invention, the BMC intelligently monitors the flow, and dynamically opens the data channel between the calculation units, so that the overload calculation unit can distribute the data to the idle unit for forwarding, thereby effectively breaking through the bandwidth limitation of the physical port of the single calculation unit, solving the problem of local flow congestion, and improving the reliability of the system. And the network throughput and the resource utilization efficiency of the whole machine system are obviously improved.

The invention provides a PCDN terminal identification and control method and device, equipment and a medium, and the method comprises the steps: obtaining a multi-dimensional behavior feature based on the communication behavior data of a target terminal, and constructing a multi-dimensional behavior feature vector based on the multi-dimensional behavior feature; inputting the multi-dimensional behavior feature vector into a pre-trained classification model to obtain an identification result of whether the target terminal is a suspicious terminal; in response to the identification result that the target terminal is a suspicious terminal, analyzing each communication flow of the target terminal in the global flow table one by one, and judging whether each communication flow is a P2P uplink flow with a content distribution intention; a bandwidth limiting measure is applied in the upstream direction of a communication flow with respect to a P2P upstream stream having a content distribution intention. According to the embodiment of the invention, the method can achieve the high-accuracy recognition of the PCDN terminal and the corresponding P2P uplink flow, and achieves the flow control with the minimum intrusiveness.

Methods and systems are provided for scaling video streams, and in particular video streams of medical data such as those from an endoscope. Video streams can be scaled to match the resolution requirements of a video display device. By performing the video scaling on a single computing chip, video scaling can be achieved while avoiding or mitigating the 8.5 Gbps bandwidth limitations of 10G networks. The use of the single computing chip can also reduce costs associated with using multiple computing chips. The result is a multi-view, cost-effective solution for displaying high-quality video streams while avoiding 10G Ethernet network bottlenecks.

The invention relates to a traffic flow prediction method and system based on space-time delay and heterogeneity modeling, and belongs to the technical field of intelligent traffic. The method aims to solve the problem of insufficient prediction precision caused by difficulty in capturing time-space propagation delay and spatial heterogeneity of traffic flow in the existing prediction technology. According to the technical scheme, the method comprises the following steps: explicitly quantifying and fusing space-time delay information by using a dynamic time warping algorithm with bandwidth limitation through a space-time dynamic delay capture module; fusing multi-granularity time periodicity and spatial characteristics through a space-time embedding module; generating dynamic weight and bias parameters of self-adaptive spatial-temporal heterogeneity through a dynamic parameter optimization module; and finally, the optimized parameters are fused into a Transform framework to carry out space-time modeling and prediction. According to the method, the complex dynamic characteristics of the traffic flow can be described more accurately, the accuracy and generalization ability of the prediction model are remarkably improved, and more reliable decision support is provided for intelligent traffic management.

The invention discloses a computing cluster task management method and a computing cluster task management device, relates to the field of computers, and realizes task scheduling which is wide in application scene, small in transmission data volume, not limited by network bandwidth and capable of reducing tail delay. The method can be applied to a computing cluster comprising a plurality of computing nodes and a shared memory, and the computing nodes are used for executing a plurality of tasks during application running and can access the shared memory. The method can comprise the steps that a first computing node in a plurality of computing nodes in a computing cluster queries global task information in a shared memory after executing a task of an application, determines a second computing node to be subjected to the task, and sends an execution request to the second computing node; and determining a to-be-executed first task by the second computing node, sending metadata of data corresponding to the first task to the first computing node, and executing the first task by the first computing node.

This invention relates to the field of network traffic rate limiting technology, and in particular to a bandwidth rate limiting method, apparatus, device, medium, and product. The method includes: after receiving an access request carrying a first floating Internet Protocol (IP) address, determining whether the first floating IP address matches any physical flow table; the output action in the physical flow table corresponds to a kernel rate limiter, and the kernel rate limiter corresponds to multiple floating IP addresses sharing the rate limit; if a match is found, determining whether the traffic of the first floating IP address exceeds the rate-limited traffic corresponding to a target kernel rate limiter, where the target kernel rate limiter is the kernel rate limiter corresponding to the output action of the matched physical flow table; if not, access is performed based on the first floating IP address. In this invention, the kernel rate limiter corresponds to multiple floating IP addresses sharing the rate limit, ensuring that multiple floating IPs share the bandwidth rate limit, improving network resource utilization and reducing costs.

The invention discloses a bandwidth limiting method and device, electronic equipment and a readable storage medium, and relates to the technical field of computers. After a product material list of a target product is obtained, a product virtual code of an object in a product material and a coding rule of the product virtual code are utilized; and analyzing to obtain characteristic codes and quantity characteristic codes corresponding to the components in the target product. Then, the characteristic codes and the number characteristic codes are used as headers, the lower portions of the headers are corresponding specification levels, and the interface bandwidth limiting tool table of the target product can be automatically generated. And importing the interface bandwidth limiting tool table into the limiting relation table library, so that bandwidth limiting based on the interface bandwidth limiting tool table can be realized. Namely, by automatically generating the product interface bandwidth limitation tool table, the PCIE resource limitation in the product BOM is automatically verified, limitation statements are not needed for limitation, the BOM creation efficiency is greatly improved, and the BOM creation accuracy is also ensured.

The invention provides a dynamic resource isolation method and device, electronic equipment and a storage medium, and relates to the technical field of communication. A resource allocation strategy containing a private network slice minimum bandwidth reservation proportion and a public network user maximum bandwidth limit proportion can be generated based on the service type identification result and the service flow prediction data; and base station scheduling parameters and core network flow routing can be adjusted according to the strategy so as to realize public and private network user resource isolation through a dynamic bandwidth allocation mechanism and network slice mapping, and meanwhile, network state changes are evaluated regularly and a resource allocation strategy is adjusted based on an evaluation result. The problem that in the prior art, resource allocation cannot adapt to dynamic service requirements can be solved.

The application relates to the technical field of cloud computing, and provides a containerized PXE service deployment system, method, device and storage medium. The system comprises: a control end configured to provide an operation page for a user to define configuration parameters; a PXE service end comprising a plurality of service containers; a PXE service configuration module configured to configure the service containers of the PXE service end according to the configuration parameters; and a PXE service scheduling module configured to deploy the PXE service end on a corresponding container host and monitor performance indexes of the PXE service end to scale the PXE service end. The system provided by the application can realize containerized distributed deployment of the PXE service end, solve the IO bottleneck and bandwidth limitation problems of the PXE service end under a large number of high concurrency conditions, and simultaneously, based on the containerized deployment of the PXE service end, realize dynamic scheduling and deployment of the PXE service end, reduce the difficulty in distributed deployment management of the PXE service end, and reduce resource waste.

The embodiment of the invention provides a multiprocessor device, a multiprocessor switching method and an intelligent lock, through integration of a first processor, a second processor, a storage medium and a switch unit, the first processor sends a notification signal to the second processor after responding to a first service request; the second processor controls the switch unit to be switched to the first conduction state, so that the first processor and the storage medium are conducted, and first data are written in. After writing is completed, the first processor informs the second processor again, the second processor controls the switch unit to be switched to the second conduction state, the second processor and the storage medium are conducted, and then the first data are read, so that the problem of low data transmission efficiency caused by bandwidth limitation of a serial port between processors in a traditional scheme is solved, and the data transmission efficiency is improved. The first processor and the second processor can use the storage medium as a data transfer station through switching of the switch unit to realize rapid transmission of data, so that the data updating efficiency is improved.

The invention discloses a coplanar waveguide feed-based T-shaped ultra-wideband transparent antenna and a design method thereof, and belongs to the technical field of ultra-wideband transparent antennae, the coplanar waveguide feed-based T-shaped ultra-wideband transparent antenna comprises a transparent dielectric substrate, and a coplanar waveguide feed structure, a radiation unit and a grounding plate structure are integrated on the same surface of the transparent dielectric substrate; the radiation unit is a T-shaped composite radiation patch; the grounding plate structure comprises two L-shaped windows, and the L-shaped windows are formed in the large-area grounding plane; and the two L-shaped windows are symmetrically distributed about the central longitudinal axis of the antenna and form partial projection overlapping with the T-shaped composite radiation patch. Through the collaborative design of the specific T-shaped composite radiation patch and the symmetrical L-shaped window, a plurality of resonant modes are effectively excited and deeply fused, so that the bandwidth limitation of the transparent antenna is overcome, and a key component is provided for a transparent integrated communication terminal; the whole structure is compact, the transparency is high, and stable and efficient ultra-wideband work can be achieved within the frequency band ranging from 4.7 GHz to 7.45 GHz.

The utility model provides a connector, an electronic assembly and electronic equipment, the connector comprises a fixing structure and a terminal, and the terminal is fixed on the fixing structure. The terminal comprises an electric connecting part and an extending part, the electric connecting part is of a metal structure, and the extending part is of an insulating structure. The electric connection part comprises a first end and a second end, the first end is fixedly connected with the fixing structure, and the second end is fixedly connected with the extension part. The terminal comprises a contact located on the electric connection portion, and the distance between the contact and the end face of the second end is smaller than or equal to 0.8 mm. According to the connector provided by the utility model, the extension part plays a role in guiding and pre-pressing, and meanwhile, the limitation on the transmission rate and bandwidth of signals is less, so that the signal transmission rate and bandwidth of the connector can be improved, and the connector is particularly suitable for transmitting high-speed signals and / or high-frequency signals. In addition, the scheme has low requirements on the preparation precision of the terminal and the connector, and the connector is simple in structure, so that the cost of the connector can be reduced.

A storage system includes at least one oversubscription-protected storage group and at least one non-protected storage group. Some or all of the primary path ports of the non-protected storage group are secondary path ports for the protected storage group. The secondary port paths are normally inactive for the protected storage group. When aggregate bandwidth on the primary port paths of the protected storage group reaches a threshold condition or is forecasted to reach the threshold condition, the secondary port paths are activated for the protected storage group. While the secondary port paths are activated for the protected storage group, both the protected storage group and the non-protected storage group are accessible via the shared ports. Access to the protected storage group and the non-protected storage group via the shared ports may be bandwidth-limited such that sharing in a predetermined ratio is ensured.

The invention discloses a graphics card online health assessment and scheduling method and device, electronic equipment and a storage medium, and relates to the technical field of computer hardware fault processing and automatic control. According to the method, health scores are calculated based on error event counting, average access delay, temperature, power consumption and access utilization rate of video memory particles, and the health score of each video memory particle is mapped into a corresponding health state grade; the video memory particles in the excellent state execute the current scheduling strategy and serve as priority target particles for task migration; starting an enhanced monitoring mode for the video memory particles in a good state, and increasing the data acquisition frequency; executing a preventive scheduling optimization strategy on the video memory particles in the warning state; and executing a scheduling strategy including task migration, bandwidth limitation, frequency adjustment and memory bank level loss on the video memory particles in the risk state. According to the method, the health state of the video memory is accurately quantified, differentiated scheduling is implemented based on the health state level, the system reliability is remarkably improved, and the downtime risk is reduced.

The invention discloses a high-speed acquisition and analysis device for an optical fiber current and voltage transformer, and belongs to the technical field of signal acquisition and transmission of optical fiber current and voltage sensors. The device comprises a photoelectric conversion module used for converting a collected optical signal of an optical fiber current and voltage transformer into an electric signal; the filtering module is used for removing high-frequency noise in the electric signal; the amplifying module is used for amplifying the electric signal which is output by the filtering module and is subjected to high-frequency noise removal; the A / D conversion module is used for converting the amplified electric signal into a digital signal; the data coding module is used for coding the digital signal; and the network port transmission module is used for transmitting the coded digital signal to a computer end through the Ethernet by utilizing a network port. According to the device, high-speed and real-time transmission of data can be realized, and the problem of data loss caused by serial port bandwidth limitation in the prior art is solved. The network port transmission can ensure that the data acquired by the FPGA (field programmable gate array) is synchronous with the sampling rate, so that the complete transmission of the data is realized.

The invention relates to a check point loading method and device in distributed training, a storage medium and a program product. The method comprises the following steps: segmenting a check point file in the shared storage device into multiple pieces of block data; for any training node in a training node cluster corresponding to the target distributed training task, loading block data corresponding to the training node from a shared storage device, and storing the block data corresponding to the training node into a local storage device of the training node; and exchanging block data among different training nodes of the training node cluster by adopting a point-to-point technology until each training node of the training node cluster respectively obtains each piece of block data corresponding to the check point file. According to the method and the device, the loading speed of the check point can be obviously increased, and the loading bottleneck problem caused by storage bandwidth limitation is effectively avoided.

The invention relates to the technical field of artificial intelligence model transmission deployment, in particular to a low-cost large-model all-in-one machine OTA method and system, and the method comprises the following steps: a local model management system defines a model version number according to a rule, and a model resource monitoring assembly periodically initiates file information query of a specified directory; the terminal model distribution component receives the model downloading task and the upgrading notification, and executes upgrading operation according to the script; the method has the beneficial effects that the cost of the server and the network channel is reduced based on the free characteristic of the public cloud disk service; the block encryption, incremental updating and multi-cloud-disk parallel downloading technology is adopted, the single-channel bandwidth limitation is broken through, the downloading time of a large model file is shortened, and the transmission efficiency is improved; in combination with a breakpoint resume technology, repeated transmission resource waste caused by network interruption is reduced; a professional team does not need to maintain a server architecture, automatic model distribution and updating are realized through a cloud disk SDK standardized interface, and the operation and maintenance difficulty is reduced.

The invention relates to the field of network control systems, in particular to a performance evaluation method for a bandwidth limitation network control system under hybrid attacks, which is realized based on a two-degree-of-freedom controller and comprises the following steps of: modeling denial of service attack and spoofing attack by utilizing a Bernoulli random process; obtaining a mathematical model of the bandwidth limitation network control system under the mixed attack; substituting the mathematical model into the tracking error signal to obtain a tracking error signal of the bandwidth limiting network control system under the hybrid attack; and obtaining a tracking error performance limit expression of the bandwidth limiting network control system under the hybrid attack according to the tracking error signal and the tracking error performance index of the bandwidth limiting network control system. According to the method, the optimal tracking performance of the system is obtained by using tools such as all-pass decomposition, partial fraction decomposition, dual Bezout equations, H2 space decomposition technologies and Youla parameterization of a controller, and the characteristics of the system and the influence of network attacks on the performance of the control system are quantitatively revealed.

A kind of avionics network virtual link configuration method based on dynamic programming and top-down, detect IP fragmentation caused by MTU, network bandwidth limit and the jitter limit caused by multiple virtual link propagation, the bandwidth calculation problem of virtual link is modeled into the optimization problem of minimization;According to the limit condition, the initial parameters of virtual link that make the total bandwidth of network minimum are obtained, the message allocation method based on dynamic programming is used when the size of message is small, and the traversal type top-down message allocation method is used when the size of message is large, to obtain the corresponding virtual link and configuration parameters when each message flow is transmitted.The message allocation method with the target of improving bandwidth margin using dynamic programming idea and top-down structure respectively can significantly improve bandwidth margin to adapt to larger scale actual data, and has fast running speed.

The embodiment of the invention discloses a vehicle flashing control method, computer equipment and a computer storage medium. According to the embodiment of the invention, the data body of the flash file is not transmitted by using the communication service of a communication protocol, but the storage position information and the related control instruction of the flash file are transmitted by using the communication service, so that the vehicle control device is guided to autonomously acquire the file data from the specified high-speed storage position. Through the separated design of a control instruction and data transmission, an acquisition path of a flash file is sent to a vehicle control device in an instruction form, so that the vehicle control device autonomously pulls the file from a high-speed external storage medium or an internal domain controller cache, bandwidth limitation and storage pressure when a traditional diagnosis bus transmits a large file are thoroughly avoided, and the reliability of the diagnosis bus is improved. The flashing efficiency is greatly improved; and the dependence on hardware resources is reduced.

A method for monitoring an aircraft's onboard communication network, comprising, for a data stream transmitted or received by a network switch, a flow limiting function (to enforce a bandwidth limit) and a security function (202) comprising: obtaining measurements of at least one stream parameter; based on the measurements obtained, calculating a value of at least one bandwidth consumption parameter, contributing to the definition of a measured bandwidth consumption profile; comparing the measured profile with a predetermined profile by comparing the calculated value with a reference value; and, if at least one difference criterion is met, performing at least one network security action. This allows the detection of an unusual (although remaining below the bandwidth limit) bandwidth consumption profile that could constitute a malicious act. Figure to be published with the abbreviation: Fig. 2

This utility model discloses an 1188B communication module and communication system, relating to the field of communication technology. The 1188B communication module includes a protocol conversion chip and a connector. The protocol conversion chip is connected to the connector via a PCIe interface and an 1188B interface. The protocol conversion chip is used to: receive Transaction Layer (TLP) data packets from the connector via the PCIe interface; convert the TLP data packets into 1188B link layer data frames; and send the 1188B link layer data frames to the connector via the 1188B interface. Therefore, the 1188B communication module provided in this application avoids the conversion scheme using 1000 Mbps Ethernet as the conversion medium, achieving seamless conversion between the PCIe protocol and the 1188B protocol through the protocol conversion chip, thereby breaking through the 1Gbps bandwidth limitation and meeting the needs of high-speed data acquisition and real-time control.

The invention discloses a terahertz signal generation system and method based on a space radio-over-fiber technology. The terahertz signal generation system comprises an optical comb generation module, an optical carrier frequency selection and modulation module, a transmission module and a photoelectric conversion module which are arranged in sequence. The optical comb generation module is used for generating an optical frequency comb; the optical carrier frequency selection and modulation module is used for screening out at least two optical carriers from the optical frequency comb and carrying out radio frequency signal modulation on the optical carriers to obtain multiple paths of optical signals; according to the invention, generation and transmission of ultra-wideband terahertz signals are realized by using a space radio-over-fiber technology, the problems of bandwidth limitation of a traditional electronic technology and overlarge loss of the terahertz signals during long-distance transmission are solved, and spurious signals generated in an optical domain signal processing process are suppressed by using an optical injection locking technology. Reconfigurable generation of terahertz signals can be realized through combined regulation and control of local oscillator signals and double lasers, and generation and transmission of high-speed, stable and low-stray broadband terahertz communication signals are realized.