117 results about "Virtual queue" patented technology

The invention discloses a fine-grained QoS isolation method and device for NVMe-oF storage, relates to the technical field of data transmission, and mainly aims to solve the problem of tail delay amplification caused by cross-tenant I / O interference in a multi-tenant environment. The method comprises the following steps: acquiring system telemetry data and a multi-tenant I / O request stream in a multi-tenant scene; performing micro-slice processing on the multi-tenant I / O request stream, and importing each obtained micro-slice into a corresponding virtual queue; determining a two-dimensional credit vector for controlling each virtual queue based on the system telemetry data; and on the basis of the two-dimensional credit vector of each virtual queue, performing scheduling and gating release on the micro-slices in each virtual queue on a preset time grid. According to the invention, the micro burst is effectively inhibited at the RNIC near end, the delay amplification path is cut off, the I / O tail delay is reduced, and the service quality predictability in the shared storage environment is improved.

The invention discloses a joint optimization method for task unloading and power control in a satellite edge computing network, and provides a time delay-energy consumption balance optimization framework oriented to dynamic task arrival aiming at the long-term random optimization problem in satellite edge computing task unloading. Through fusion of Lyapunov optimization and a virtual queue mechanism, the long-term delay constraint is successfully converted into a time slot level optimization target which can be executed online, and a self-adaptive decision process is realized. And aiming at the generated non-convex optimization problem, constructing a robust solution algorithm under a PDD framework, and integrating secondary transformation and an SCA technology to realize efficient solution. Simulation experiments show that the method can obviously reduce the energy consumption in a random task arrival scene, and strictly meets the delay constraint requirement at the same time.

The invention belongs to the related technical field of wireless communication, and discloses a resource joint scheduling method and system oriented to an edge cloud fusion industrial internet, and a server. The method comprises the steps that a scheduling model in a long-time scheduling period of a current edge server is constructed, the long-time scheduling model comprises T discrete time slot sub-models, the optimization objective of each time slot sub-model is to minimize an objective function under a corresponding time slot, and the objective function comprises an equation # imgabs0 # Q (t) which is the length of an energy consumption virtual queue in a time slot t, the queue of each time slot meets the condition that # imgabs1 # t = 0,..., T-1; zeta (t) is the time delay satisfaction rate of all calculation task requests in the time slot t; and converting all the time slot sub-models into a Markov decision process, and performing optimization by adopting a deep reinforcement learning algorithm to obtain a memory, computing resource and bandwidth resource allocation strategy of each time slot in a long-time scheduling period. According to the method, a long-period resource allocation decision can be made, so that the time delay and the energy consumption of the industrial Internet of Things in a long period can be considered at the same time.

The invention relates to a multi-energy information physical system scheduling method, system and device with bidirectional aging perception, and belongs to the technical field of telecommunication. The method comprises the following steps: firstly, constructing a physical domain multi-energy system model and an information domain multi-energy bidirectional timeliness model based on an information physical fusion multi-energy system; secondly, constructing an information physical fusion multi-energy system scheduling model considering the multi-energy bidirectional timeliness to minimize the global scheduling cost of the system; bidirectional timeliness constraint is converted through a virtual queue, and decoupling of short-term decision and long-term constraint is achieved. On the basis, model learning at an offline stage and extrinsic matching optimization at an online stage are combined, an information interaction sequence and a consistency variable weight are dynamically adjusted, information domain interaction and physical domain regulation and control are coordinated, and through distributed iterative optimization, the calculation complexity is effectively reduced, and the calculation efficiency is improved. The convergence speed of the consistency variable and the system regulation and control efficiency and stability are improved, and the system scheduling cost is reduced.

The invention belongs to the field of metallurgical industry automation, and relates to an iron and steel joint enterprise energy management system scheduling method based on Lyapunov optimization and deep reinforcement learning. In order to solve the problems that an existing system depends on static rules or experience, is difficult to adapt to dynamic changes, is low in energy scheduling efficiency and the like, a multi-dimensional state space covering energy storage, a virtual queue and environment parameters is constructed, and a scheduling action space including capacity adjustment, multi-energy medium charging and discharging control and the like is defined; a Lyapunov drift function is introduced as a stability constraint, a reinforcement learning agent of an Actor-Critic framework is constructed, and global energy collaborative optimization is realized in combination with an energy sub-optimization problem solving and parallel sampling mechanism. The method can give consideration to scheduling economy and system stability, has dynamic adaptive ability, ensures project feasibility of the scheme, can improve energy utilization efficiency, reduces cost and carbon emission, and is suitable for iron and steel enterprises of different scales.

The invention discloses a virtual power plant multi-target time domain coupling feasible region construction method and system, and the method comprises the steps: building a neighborhood feasible region vertex search model of a virtual power plant, solving the neighborhood feasible region vertex search model through a vertex search method, and obtaining a time domain coupling feasible region at a t-1 moment and a t moment; according to the method, the wide-time-domain output of the virtual power plant is stably described as that the net flow value of the virtual queue is zero in a certain time through the virtual queue, an optimization objective function of a neighborhood feasible region vertex search model comprises a Lyapunov function representing the crowdedness degree of the virtual queue, and the crowdedness degree of the virtual queue is calculated by solving the optimization objective. And a time domain coupling feasible region considering the stability of the overall virtual queue of the virtual power plant can be obtained. The method can improve the flexibility of the power system, optimize the energy structure, save energy and reduce consumption.

The invention discloses an early gastric cancer intelligent screening and risk layering method and system based on multi-modal data fusion and deep learning. The method comprises the following steps: acquiring a gastroscope image, a pathological image, a medical image and clinical data of a patient, performing standardization processing and forming unified feature representation; training the multi-branch neural network model group, and generating a virtual queue tag based on each mode; fusing a multi-modal queue division result to generate a patient comprehensive queue label; adaptively selecting a decision path according to the label, and executing feature fusion and analysis of corresponding depth; generating a structured report containing gastric cancer risk levels and individualized monitoring suggestions; and deploying the system and dynamically tracking the change of a patient queue to realize continuous monitoring of disease progression. Cross-modal semantic alignment, uncertainty evaluation and a Markov prediction model are introduced into the system, the screening accuracy, credibility and dynamic risk evaluation capability are improved, and the system is suitable for early screening and personalized management of gastric cancer.

A method for handling data packets by a communication node in a communication network, the method comprising storing received data packets in at least two physical queues, wherein a first of said at least two physical queues is associated with low latency data packets and a second of said at least two physical queues is associated with high latency data packets, wherein each data packet is stored in one of the at least two physical queues based on a delay characteristic associated with the data packet, for each received data packet, storing an associated information record in at least two virtual queues, VQs, wherein associated information for data packets stored in said high latency physical queue is stored in a second of said at least two virtual queues and wherein associated information for data packets stored in said low latency physical queue is stored in both said first and second of said at least two virtual queues, serving data packets from the at least two physical queues, using at least two Congestion Threshold Values, CTVs, wherein a first of said at least two CTVs is applicable to data packets in said low latency physical queue and wherein both said first and second of said at least two CTVs are applicable to data packets in said low latency physical queue and data packets in said high latency physical queue, wherein said at least two CTVs are used for at least one of dropping and marking packets based on their associated information.

Aspects presented herein relate to methods and devices for display processing including an apparatus, e.g., a DPU. The apparatus may configure a virtualization component for a set of workloads associated with a set of VMs. The apparatus may also execute the set of workloads associated with the set of VMs for a set of virtual queues. The apparatus may also perform at least one of: (1) a mapping of the set of VMs to the set of virtual queues, (2) a security check for an ability of each of the set of VMs to access a set of software registers, or (3) a mapping of each of the set of virtual queues to each of a set of control data paths.

The invention discloses a multi-time-scale energy storage peak regulation frequency modulation time division multiplexing method, which comprises the steps of dynamically dividing a peak regulation frequency modulation working area day ahead based on an energy storage cluster available capacity and a load curve; peak valley arbitrage, peak regulation compensation and frequency modulation benefits are considered, an energy storage cluster peak regulation and frequency modulation optimization model is established by taking maximization of the economic benefits of energy storage cluster peak regulation and frequency modulation as a target, and the overall optimal charging and discharging power of the energy storage cluster is obtained; based on this, the differential adjustment requirements of long-time scale peak regulation and short-time scale frequency modulation are considered, and the power constraint, capacity constraint, charging and discharging power range constraint and virtual queue backlog variable constraint capable of tracking the energy storage capacity change of each energy storage power station in the energy storage cluster are obtained. And constructing a peak regulation and frequency modulation distribution model of the energy storage cluster meeting the constraint condition. According to the invention, the performance of the energy storage cluster in the peak regulation and frequency modulation task is ensured not to be lost while multi-task coordination is realized, and the utilization rate of the energy storage power station is improved.

The invention discloses a parameter adjustment method and system for a Wi-Fi module. The method comprises the following steps: acquiring a multi-mode interference sample; constructing a spatio-temporal annotation data set; constructing an LSTM-attention model to carry out interference prediction, and outputting a prediction result; according to the real-time IQ sampling data, a prediction result and a historical interference database, constructing a dynamic interference tensor, and through confidence coefficient weighted CP decomposition and gating feature fusion, outputting a fused enhanced feature; a QoS vector is dynamically generated according to the equipment type and the service demand, quality evaluation is carried out on the service through the final quality value, and the service quality is dynamically optimized; according to the final quality value, the prediction result and the current radio frequency parameter, a virtual queue is constructed and updated, the quality deviation condition is monitored, and the radio frequency parameter is dynamically adjusted. According to the method, through closed-loop cooperation of interference modeling, service awareness and an optimization framework, the self-adaption problem of the Wi-Fi module in a high-dynamic interference environment is solved, and the method is more suitable for complex scenes such as an industrial wireless control system.

The invention discloses a data issuing method, a data uploading method and a related device, and relates to the technical field of data processing.The data issuing method comprises the steps that when a network issuing instruction is received, a first direct storage access request is initiated according to a first issuing descriptor, and a first issuing data packet is obtained from a storage area, transmitting the first issuing data packet to the first equipment through the first issuing virtual queue; and when a calculation acceleration issuing instruction is received, initiating a second direct storage access request according to a second issuing descriptor, acquiring a second issuing data packet from the storage area, storing the second issuing data packet into the cache, and transmitting the second issuing data packet to a target processing module for processing according to a protocol frame type of the second issuing data packet, and the response frame is generated and returned to the upper computer, so that the technical problem that a heterogeneous accelerator is complicated in design when multifunctional data transmission is realized in related technologies is solved, and data transmission of a network function and a calculation acceleration function under the management of the same virtual queue is realized.

A traceable method for task replication and resource allocation of a large-scale machine type communication network belongs to the technical field of wireless communication and mobile edge computing, and comprises the following steps: based on a Lyapunov optimization theory, establishing physical and virtual queue quantization constraints, and converting long-term optimization into a drift and penalty minimization problem of each time slot; decomposing into task replication, resource block allocation and edge server queue management sub-problems, and sequentially solving; determining an optimal task replication scheme, and storing the optimal task replication scheme as a first decision-making voucher; determining a resource block allocation scheme, and storing the resource block allocation scheme as a second decision-making voucher; determining an optimal edge server queue management scheme which comprises an edge server task processing rate and a task discarding amount, and storing the optimal edge server queue management scheme as a third decision-making voucher; updating a queue state, packaging the decision-making voucher and the queue state into a periodic transaction, writing the periodic transaction into a block chain, and forming an operation auditing log; according to the method, the optimal balance of long-term operation cost is realized, and the decision-making process has traceability.

The invention discloses a dynamic state updating scheduling method in a multi-terminal wireless communication system. The method comprises the following steps of: aiming at the problem that uplink resources are limited, constructing a virtual queue to represent the historical resource use condition of a terminal; dynamic updating indexes are designed to quantify urgency of state updating; high-value terminal screening and an optimal scheduling decision are executed in each time slot; and a Lyapunov optimization framework is creatively combined with a dynamic threshold mechanism. By adopting the method provided by the invention, the average information urgency can be minimized, and the information transmission performance of the system is effectively improved.

The application discloses a data issuing method, an uploading method and related devices, and relates to the technical field of data processing. The data issuing method comprises the following steps: when a network issuing instruction is received, a first direct storage access request is initiated according to a first issuing descriptor, first issuing data packets are acquired from a storage area, and the first issuing data packets are transmitted to a first device through a first issuing virtual queue; when a computing acceleration issuing instruction is received, a second direct storage access request is initiated according to a second issuing descriptor, second issuing data packets are acquired from the storage area, the second issuing data packets are stored in a cache, the second issuing data packets are transmitted to a target processing module for processing according to a protocol frame type of the second issuing data packets, a response frame is generated, and the response frame is returned to an upper computer, so that the technical problem that a heterogeneous accelerator design is complex when multifunctional data transmission is implemented in the prior art is solved, and data transmission of network functions and computing acceleration functions under the management of a same virtual queue is implemented.

The application relates to the technical field of network slices, in particular to a message priority scheduling test system for high-density V2X communication, which comprises a V2X slice congestion state monitoring module, which is used for monitoring the C-V2X message arrival rate and the physical buffer occupancy rate of each network slice, calculating a virtual queue length value, and aggregating the virtual queue length value of all network slices with corresponding slice identifiers. In the application, the C-V2X message arrival rate and the physical buffer occupancy rate of each network slice are synchronously observed and numerically integrated, a slice congestion state representation system with time sensitivity and dynamic adaptability is established, the identification capability for network operation pressure is improved, and through coupling modeling of the message life cycle and the slice bearing state, a potential hazard degree index for a single message is constructed, so that the scheduling system has stronger perceptual granularity and task level discrimination capability.

The application provides a virtual power plant dynamic feasible region solving method and system based on coupling constraint decoupling, belongs to the technical field of power plant resource allocation management, establishes a virtual power plant model considering distributed energy and flexible load resources, and determines a virtual power plant feasible region solving model based on a vertex search method; through Lyapunov optimization, virtual queues of energy storage, heating ventilation air conditioning and electric vehicles and other flexible loads are respectively established, and the time coupling constraint of the flexible load is decoupled into a virtual queue stability problem; a virtual time queue is constructed to integrate the decoupled optimization problem into the virtual power plant dynamic feasible region solving of day-ahead and day-ahead, so that the day-ahead and day-ahead feasible region solving algorithm is obtained. Based on the Lyapunov optimization theory, the time coupling constraint is decoupled, a weight coefficient is introduced, and a mathematical model taking the vertex search and queue stability as targets is provided, and finally the virtual power plant dynamic feasible region solving of day-ahead and day-ahead is realized.

The application discloses an edge resource allocation method and device based on long-term value fairness and overlapping coalition game, a medium and equipment, and belongs to the technical field of mobile edge computing and resource scheduling. The method builds a distributed crowdsourcing system architecture, establishes an ESP multi-dimensional resource capability, a dynamic task flow and a reputation evaluation model, models the resource collaboration problem as an overlapping coalition game, builds a value loss virtual queue, converts it into a single time slot optimization problem by using a Lyapunov drift plus penalty framework, and designs a distributed potential game algorithm guided by rights and interests to realize the collaborative optimization of system efficiency and long-term fairness.

The invention discloses a network virtualization resource management method and system, and relates to the technical field of intelligent resource management, and the method comprises the steps: initializing an AoL virtual queue, dynamically updating the queue through a Lyapunov drift method, dividing scene feature groups in combination with an ODSS algorithm, fusing prediction results of multiple models, correcting deviation through a Lyapunov-ODSS objective function, and refining the prediction results according to priorities. And an OPT-JR multi-objective function is defined, and an optimal resource allocation vector is solved by using an NSGA-II algorithm. According to the method, the Lyapunov drift theory is introduced to carry out resource regulation and control, the ODSS feature screening algorithm is combined to carry out feature selection, the robustness and real-time performance of resource management are improved, multiple models are fused to carry out collaborative prediction, the precision and stability of resource demand prediction are improved, and the utilization efficiency and intelligent scheduling of resources in a virtualized network are realized.

The application discloses an edge intelligent body flexible scheduling method based on Lyapunov optimization and Stackelberg game. The application constructs a cloud-edge-end three-layer edge intelligent body federated learning network architecture to complete distributed collaborative training under the premise of protecting data privacy. In view of the problems of heterogeneous edge intelligent body network nodes, non-independent and identically distributed data and limited communication resources, a long-term joint optimization model of client selection and bandwidth allocation is established, and model accuracy, time delay and energy consumption are considered. Lyapunov optimization is used to decouple long-term random optimization into real-time decision-making per time slot, and to convert constraints into virtual queue stability control. Dynamic client selection is realized based on Stackelberg game, and high-quality nodes are adaptively selected. Adaptive bandwidth allocation is completed through an evolutionary algorithm, and communication requirements of key clients are preferentially guaranteed. The application can significantly improve model accuracy, reduce time delay and energy consumption, guarantee long-term stability of the system, and is suitable for high-dynamic edge intelligent body federated learning scenes.

The present application relates to the field of distributed large model inference, and particularly relates to a distributed large model inference method based on token perception and Lyapunov optimization. The scheme comprises: after receiving a user input prompt, predicting the output token length of the corresponding task through a length perception semantic module, calculating the corresponding total amount of work, based on the token length prediction result, the real-time state of the device and the long-term virtual queue state, generating a final unloading decision through the Lyapunov guided unloading optimization module iterative unloading algorithm; the client unloads the task to the corresponding cloud edge device according to the final unloading decision, and the device executes the large model inference; the scheduler updates the virtual queue length according to the actual task execution time and the device computing power utilization rate, and provides input for the next time slot decision. The present application is suitable for inference systems utilizing distributed large models.

Location-based virtual queuing of vehicles is described. A request to join a virtual queue may be received from a user application. In response to receiving the request, the server may determine, using a geo-fencing feature, if the user computing device running the user application is within a first predetermined distance of a queue location, where the queue location is a destination of the requesting user vehicle. Then, in response to receiving confirmation that the requesting vehicle user is within the first predetermined distance, the server may automatically create and add a reservation message to a data structure that includes a plurality of reservation messages. Also, in response to the creating the reservation message, the server may provide the requesting user application a notification with the position within a queue managed by the data structure and an estimated wait time that is generated using information stored within the data structure.

Presented are cryptographic digital assets for retail products, methods for making / using such cryptographic digital assets, and computing systems for generating, intermingling, and exchanging blockchain-protected products. A method for provisioning cryptographic digital assets associated with retail product transfers includes broadcasting notifications of a future transaction of a retail product, and receiving, over a distributed computing network from the computing devices of multiple users, requests to participate in the transaction. A select number of users is added to a virtual line associated with the retail product transaction; from the virtual line, a first user is selected to receive the retail product and a second user is selected to receive a cryptographic digital asset containing a digital retail product and a unique digital asset code. The cryptographic digital asset is transferred to the second user's digital wallet, and the unique digital asset code is recorded on a record block of a blockchain ledger.

The application discloses a kind of based on equipment training time fair federated learning method, mainly including the equipment selection method for guaranteeing training time fairness from long-term perspective, and the model aggregation method cooperating with equipment selection method.Federated server selects equipment by means of virtual queue and predicted equipment training time, and selects equipment to participate in each round of training;Selected equipment downloads global model from federated server, trains on local data, and sends back model to federated server after training;After receiving all the returned models, the federated server uses the model aggregation method to aggregate the models;The above process is repeated until the federated learning task is completed.In addition, with the help of training time fairness constraint, the application ensures that all devices can participate in training, while preventing excessive training time.The model aggregation method cooperates with the equipment selection method to increase the weight of the devices that are selected less frequently during aggregation, allowing each device to contribute to the model.

The invention discloses a cross-regional construction site-oriented edge smart center data scheduling method and system, and the method comprises the following steps: receiving construction site heterogeneous data streams, and outputting a hierarchical data stream set; constructing a virtual queue to represent backlog and default risks, and adopting an online optimization mechanism of an SODA algorithm to generate a scheduling strategy; queueing and channelizing the hierarchical data stream set according to a scheduling strategy to obtain a channelized data stream set; based on the channelized data stream set, dynamic degradation is executed according to a scheduling strategy when resources are insufficient, and a degraded data stream set is generated; the switching cost of the degraded data stream set is calculated, and a stable scheduling strategy is obtained; freezing current scheduling execution according to a stable scheduling strategy, and outputting a supplementary transmission task sequence; and performing duplicate removal and selective supplementary transmission on the supplementary transmission task sequence, and outputting an audit record through an auditable replay mechanism. According to the invention, the timeliness guarantee capability and the data consistency of cross-regional data transmission are improved.

The application discloses a fine-grained QoS isolation method and device for NVMe-oF storage, relates to the technical field of data transmission, and mainly solves the problem of tail delay amplification caused by cross-tenant I / O interference in a multi-tenant environment, and the method comprises the following steps: acquiring system telemetry data and multi-tenant I / O request flow in a multi-tenant scenario; performing micro-slicing processing on the multi-tenant I / O request flow, and importing each micro-slice obtained into a corresponding virtual queue; determining a two-dimensional credit vector for controlling each virtual queue based on the system telemetry data; and scheduling and gating the micro-slices in each virtual queue on a preset time grid based on the two-dimensional credit vector of each virtual queue. The application effectively suppresses micro-bursts in the RNIC near end, cuts off the delay amplification path, reduces I / O tail delay, and improves the service quality predictability in a shared storage environment.

An interface module connected to a storage controller and relaying communication between a host and the storage controller. The host and the storage controller use different communication protocols, include computing cores and a memory. At least one of the computing cores is a IO core in communication with the host, each of the IO cores has a virtual queue in the memory and the IO core is configured to execute a reception process of converting communication data including a command to the storage controller outputted by the host into a storage interface command and store the communication data in the virtual queue. The IO core executes a sending process for sending to the host by converting to communication data takes out a storage interface command from the virtual queue, and in same session, same IO core performs the reception process and the transmission process.

The invention relates to an optimization method for information age and energy consumption in a hybrid industrial wireless network, and belongs to the technical field of industrial Internet of Things. Aiming at the problems of low scheduling efficiency, high periodic data overdue rate, insufficient aperiodic data throughput and unbalanced energy consumption under a dynamic channel condition, a joint optimization scheme based on Lyapunov optimization and a double-end Actor network is provided. By constructing an overdue rate and throughput virtual queue, converting long-term constraint into a real-time optimization problem, establishing a Markov decision model, adopting a double-end Actor network to decouple and update scheduling and channel allocation decision, and combining a conflict detection mechanism and dynamic weight adjustment, the periodic data overdue rate is reduced by 40-60%, and the real-time performance of the system is improved. And the aperiodic data throughput is improved by more than 25%. The method solves the problem of collaborative optimization of information freshness and energy consumption in a mixed service scene, and is suitable for real-time control and monitoring scenes of the industrial Internet of Things.