15results about How to "Guaranteed Latency" patented technology

The invention relates to the technical field of intelligent Internet of Vehicles, in particular to a task unloading method in a vehicle-mounted edge computing environment based on deep reinforcement learning, which comprises the following steps: acquiring a task vehicle to generate a to-be-executed task, uploading the task to an RSU, generating a scheduling decision according to the current state of the task and available resources, and sending the scheduling decision to the RSU. The task is allocated to a service vehicle or locally executed; performing priority division on the tasks by utilizing an analytic hierarchy process, and obtaining the current state of each to-be-executed task; obtaining service vehicle numbers and task vehicle numbers of the tasks based on task states and priority level utilization model scheduling, and calculating unloading and scheduling strategies; obtaining available computing resources of the current service vehicle; and training the sequence to the sequence model by using an Actor-Cr-it ic algorithm to obtain an optimal task part unloading and scheduling strategy. The computing resources of the service vehicle and the edge server can be fully utilized, the execution time delay of all tasks in the whole time period is smaller, and the task execution success rate is higher.

The invention discloses a millimeter wave large-scale MIMO (Multiple Input Multiple Output) hybrid precoding design method for URLLC (Uniform Resource Logical Link Control), which comprises the following steps: network side equipment adopts a hybrid precoding architecture, adopts a short packet to transmit a signal, measures the transmission performance according to a finite block length information theory; the method comprises the following steps: establishing a URLLC-oriented millimeter wave large-scale MIMO hybrid precoding optimization model based on performance measurement criteria such as maximization and rate or maximum and minimum fairness under multi-user communication business service requirements, transmission total power constraints and feasible phase set constraints of phase shifters with different resolutions; the reconstruction optimization problem enables the augmented Lagrange problem to have a decomposable structure, and the hybrid precoding optimization problem is solved through double iteration; the augmented Lagrange problem is rapidly solved in the inner layer iteration to update the original optimization variable, and the dual variable and the penalty factor are adaptively updated in the outer layer iteration. The system performance can be effectively improved, and the implementation cost and energy consumption of the system are reduced.

The application provides a method and device for quantifying elastic KV cache through computing power of an intelligent computing cloud platform, and relates to the technical fields of intelligent computing centers, intelligent computing centers, computing power infrastructure and intelligent computing cloud technology.The method comprises the following steps: S1, dividing historical tokens into multiple cache blocks, quantifying KV data and writing the data into corresponding cache blocks, and selecting multiple candidate anchor points; S2, calculating block-level summary data; S3, in response to a new target token, scoring the cache blocks to generate cache block scores and screening out candidate cache blocks; S4, calculating uncertainty index data and determining a target cache block with to-be-restored precision according to the uncertainty index data; and S5, locating an upstream target anchor point and locally playing back the historical tokens based on the upstream target anchor point to generate target high-precision KV data of the target cache block.The application can greatly improve the quantification effect of KV cache of the intelligent computing cloud platform.

The embodiment of the invention relates to the technical field of networks, and particularly provides a message processing method, device and equipment, a network card and a computer program product. The message processing method is applied to a field programmable gate array (FPGA) module, the FPGA module is in communication connection with a driving module, and the driving module is used for driving data transmission between the FPGA module and a host memory. The method comprises the following steps: the FPGA module obtains a current message to be processed through the driving module; under the condition that the current message is the message needing to be discarded, the FPGA module determines an order-preserving queue used for achieving message processing operation, and the order-preserving queue at least comprises order-preserving information of all the messages needing to be subjected to order-preserving operation; and the FPGA module discards the current message and releases the order-preserving information of the current message in the order-preserving queue. According to the embodiment of the invention, the situation that the header is blocked in the order-preserving queue due to the abnormal packet loss operation of the software is avoided, and meanwhile, the influence on the time delay and the performance of message forwarding is avoided.

The present application discloses a connection switching processing method, apparatus, related equipment and storage medium. The method includes: a terminal in a heterogeneous network generates first information based on at least the measured signal strength of a second access network device in the heterogeneous network; reporting the first information to a first access network device in the heterogeneous network; the first information is used by the first access network device to switch the terminal; the first access network device can provide the terminal with a first access mode; the second access network device can provide the terminal with a second access mode; the first mode is different from the second mode.

The present invention discloses a method and system for reducing remote control delay. A multi-path transmission channel is established between the control end and the controlled terminal. The direct channel transmits the video data stream. All channels transmit the instruction interaction message at the same time. The transmission channel, instruction sequence and execution result are distinguished by the control instruction coding and the execution response coding. Then, the optimal delay channel is determined according to the execution response message, which is used to determine the transmission channel sequence for sending the next instruction. On the basis of ensuring the direct transmission of high-bandwidth video stream, multiple transmission paths are established. With the control instruction sending and the execution response coding, instructions can be sent to all established transmission paths at the same time. The existing smart home remote control method solves the technical problems that the control response delay is large, it is difficult to maintain high-quality video monitoring effect, and the user experience is affected by establishing a direct transmission channel between the controlled terminal and the control end, and transmitting the video monitoring data stream and the controlled terminal control instruction at the same time.

The embodiments of the present invention relate to the field of network technology, and specifically provide a message processing method, device, equipment, network card and computer program product. The message processing method is applied to a field programmable gate array FPGA module, and the FPGA module is communicatively connected to a driver module, and the driver module is used to drive the data transmission between the FPGA module and the host memory; the method includes: the FPGA module obtains the current message to be processed through the driver module; when the current message is a message to be discarded, the FPGA module determines the order-preserving queue used to implement the message processing operation, wherein the order-preserving queue at least includes: the order-preserving information of each message that needs to be subjected to the order-preserving operation; the FPGA module discards the current message and releases the order-preserving information of the current message in the order-preserving queue. In this embodiment, the situation of head blocking in the order-preserving queue due to abnormal software packet loss operation is avoided, and the delay and performance affecting the message forwarding are avoided.

The invention relates to a data node management method and device, computer equipment, a computer readable storage medium and a computer program product. The method comprises the following steps: determining a plurality of namespaces of a disk; the plurality of namespaces are obtained by dividing the disk, and the plurality of namespaces are storage units with data isolation; allocating a processor core according to each namespace, and binding each namespace with the allocated corresponding processor core to obtain the processor core bound with each namespace; and according to the processor core bound with each namespace, managing the data node corresponding to each namespace. By adopting the method, the processing efficiency can be improved.

The application belongs to the technical field of vehicle communication, and discloses a vehicle fog service security communication system and method based on a blockchain and a terminal, which comprises the following steps: a trusted authority (TA) verifies the original id of a vehicle, allocates a pseudo id to the vehicle, and uses an elliptic curve digital signature algorithm to sign the registration information of the vehicle; a roadside unit (RSU) performs identity authentication and data transmission with a vehicle-mounted unit that sends a communication request; a vehicle-mounted unit (OBU) performs identity authentication and data transmission with a nearby RSU; a blockchain uses a consortium blockchain as a decentralized underlying architecture to be instantiated; the application ensures the ultra-low delay and ultra-high reliability of vehicle fog service when negotiating a security communication key, uses a fog node at the network edge, improves the reliability of the network, and ensures low delay of the network. The communication scheme of the application is lightweight and can resist various known attacks.

The present disclosure provides a method, device, headset and medium for controlling headsets. The method includes: detecting the current working mode of the headset; when it is detected that the current working mode is a game mode and the game mode is a continuous audio mode, configuring the buffer value of the buffer zone of the headset to be a first buffer value; when it is detected that the current working mode is a game mode and the game mode is a discontinuous audio mode, configuring the buffer value to be a second buffer value; wherein the first buffer value is greater than the second buffer value.

This application discloses a latency allocation method, apparatus, device, storage medium, and computer program product. The method includes: a first functional entity splitting the total latency requirement of a first task into a computation latency requirement and a transmission latency requirement; sending the computation latency requirement to a second functional entity and sending the transmission latency requirement to a network device; wherein the second functional entity is used to split the computation latency requirement into a network-side computation latency requirement and / or a terminal-side computation latency requirement; and the network device is used to split the transmission latency requirement into an uplink transmission latency requirement and a downlink transmission latency requirement.

This invention discloses an on-chip polarization-insensitive detector system. The system splits arbitrary polarized light into two branches for output (TE and TM modes), and these two branches are then output to both ends of the detector; the output power of the two branches is equal. The system includes an edge coupler, a waveguide interlayer coupler, an on-chip PRBS device, a 1-to-2 optical power splitter, and a detector. This invention's polarization-splitting, wave-splitting, and wave-combining method ensures that any input polarization state can be supplied with almost constant and identical optical power to both ends of the detector, mitigating the large-scale carrier shielding effect while maintaining polarization insensitivity. The incident polarization rotation beamsplitter and the wave-combining polarization beamsplitter use a mirror-image configuration to compensate for the time delays of the TE and TM modes, ensuring low polarization-dependent delays throughout the entire link when the polarization state changes.

The invention discloses a service flow processing method, device and equipment and a computer readable storage medium. The method comprises the steps of starting a target intrusion prevention system IPS engine in a target Pod associated with a target IP when an average value of monitoring index values of service traffic passing through a firewall in a preset duration is greater than a corresponding preset threshold value; when the target IPS engine is in a running state, a first state notification is sent to a firewall, so that the firewall stops analyzing the target service flow corresponding to the target IP after receiving the first state notification; and when the target service traffic reaches the target Pod, guiding the target service traffic to the target IPS engine, so that the target IPS engine performs intrusion prevention detection on the target service traffic. According to the method and the device, when the service flow is relatively high, the target service flow in the service flow is drained to the target IPS engine for intrusion prevention detection, transverse expansion of a firewall is not needed, and the time delay and the packet loss rate of the service can be effectively ensured to meet user requirements.

The invention discloses a Stackelberg game-based calculation unloading and cooperative calculation method in an unmanned aerial vehicle. The method comprises the steps of constructing a mobile edge calculation system; defining a task unloading strategy of the MUE and a cooperative forwarding strategy of the UAV; establishing system cost models including a G2A communication model, an A2A communication model, a calculation queuing model and a system energy consumption model, and establishing a task completion time delay model including constructing an optimization problem; modeling the optimization problem as a Stackelberg game; the UAV is used as a leader, and the MUE is used as a follower; alternately updating the cooperative forwarding strategy and the task unloading strategy by adopting a distributed iteration strategy solving algorithm, and outputting a balanced unloading strategy and a balanced forwarding strategy; and the MUE offloads the task to the UAV according to a balanced offload strategy, and controls the UAV to forward the task to an adjacent UAV or to process locally according to a balanced forwarding strategy. According to the method, cooperative processing of tasks is realized based on the Stackelberg game, and the comprehensive cost is effectively reduced.

The application relates to the field of robot production lines, and particularly relates to a digital twin simulation system based on robot action features, which comprises a data management unit and a twin simulation unit. The data management unit is connected to a robot management unit, collects original state data of a production line robot maintained by the production line robot management unit, carries out data cleaning and format conversion on the original state data to form twin state data, then carries out data packaging according to a subscription configuration, and sends the data to the twin simulation unit by using different transmission channels. The twin simulation unit obtains the twin state data after unpacking, then sends the twin state data to each twin simulation robot according to a registration configuration, updates the state, and displays in the form of an animation. The application further comprises a simulation method. The application collects data of a real robot or data provided by simulation software, solves the problem of robot simulation animation display, reduces the input of modeling personnel, and can also reduce the process requirements of non-professional personnel.