12 results about "Bandwidth reservation" patented technology

The invention relates to the technical field of computer storage, and particularly provides a large model training storage resource dynamic allocation method, device and system, and the method comprises the steps: collecting a workload index and a storage system state index of a large model training task; inputting the real-time monitoring data into a convolutional neural network (CNN) model, and outputting a feature identifier of a current training stage; inputting the real-time data and the stage identifier into a long short-term memory network LSTM model, and predicting future bandwidth, IOPS and storage space requirements; and generating a resource allocation scheme through multi-objective optimization according to a prediction result in combination with a system state, and executing load allocation, data layering and bandwidth reservation operations. According to the method, training stage perception and resource demand prediction are realized through cooperation of the CNN and the LSTM, and dynamic allocation and advanced scheduling of storage resources are realized, so that the resource utilization rate, the training efficiency and the system stability are improved.

The invention belongs to the technical field of data center network and clock synchronization, and discloses a dynamic time synchronization path optimization method and system based on topology awareness. The invention provides a closed-loop framework of state awareness-path optimization-resource assurance-dynamic adjustment and optimization: in the first stage, network topology, link load and PTP message performance data are periodically acquired through an SDN controller, and a global state view is constructed; in the second stage, the optimal path is dynamically calculated for the PTP synchronous flow through a multi-constraint optimization algorithm with the purpose of minimizing transmission delay jitter, and deterministic transmission is achieved in combination with bandwidth reservation and a priority queue; and meanwhile, a seamless switching mechanism of building first and disassembling second is introduced to adapt to the network state change in real time. On the basis of multiplexing existing hardware, the time synchronization error is remarkably reduced, the transmission stability and the dynamic adaptability are improved, microsecond-level time synchronization can be efficiently achieved, and the method is suitable for high-precision service scenes.

The invention discloses an in-vehicle network active bandwidth reservation type burst flow containing method. The method comprises the steps that 1, the injection time of ST flow to an in-vehicle network is calculated, and a gating list is configured; 2, residual bandwidth resources, RC flow available bandwidth and bandwidth reservation of any link are calculated; 3, calculating an injection time slot of RC to the vehicle-mounted network and configuring a gating list; 4, calculating the injection time of burst flow to the vehicle-mounted network and configuring a gating list; 5, the wakeup bandwidth amount of any link of the vehicle-mounted network is calculated; and 7, recalculating the injection time slot of the RC flow influenced by the burst flow to the vehicle-mounted network, and configuring a gating list. According to the method, the bandwidth is actively reserved when the ST flow and the RC flow are scheduled and registered, and the conflict flow is accommodated on the premise that the real-time transmission quality of the registered flow is not influenced, so that the real-time transmission quality requirement of the registered flow and the ultra-low delay requirement of burst flow transmission are ensured, and an intelligent driving system is enabled to operate normally so as to improve the driving safety and reliability of a vehicle.

A method of scheduling a Real Time (RT) task, includes: receiving a task; obtaining a yield time of the RT task based on one of an execution deadline of the RT task, an execution deadline of next RT task subsequent to the RT task, and a maximum execution time associated with an execution of the next RT task subsequent to the RT task; creating a bandwidth reservation task having a deadline; inserting the RT task along with the bandwidth reservation task into a RT wait queue based on the deadline of each of the RT task and the bandwidth reservation task in accordance with an Early Deadline First (EDF) criteria; and scheduling an unscheduled task based on an available-time of the RT wait queue in accordance with the EDF based scheduling.

Spectrum allocation in non-terrestrial wireless networks is efficiently and intelligently managed by predicting traffic loads for beam cells using historical network traffic data and a trained model. Beam cells with higher predicted traffic loads are assigned larger spectrum bandwidths, while other beam cells are assigned smaller, non-overlapping bandwidths that dynamically change frequencies according to a predefined hopping sequence. Example solutions disclosed herein optimize bandwidth utilization and minimize interference, enhancing service quality and coverage in satellite communication systems.

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 invention relates to the technical field of computing power scheduling, and discloses a computing power intermediate station resource layer full-link cooperative scheduling method oriented to multiple service scenes. According to the method, a resource request carrying a service scene label is received, the requirements for GPU computing power, a memory and storage IO are analyzed, and a standardized resource descriptor is generated. Meanwhile, states of all physical nodes are dynamically collected, a global resource view is constructed, resource descriptors are matched with the view, priorities are distributed in combination with service scene labels, and a resource reservation mechanism is triggered. And mapping the task to a target node according to a matching result, and executing GPU computing power division, memory allocation and network bandwidth reservation. Strategies are dynamically adjusted in the scheduling process, and efficient utilization of resources is ensured.

A method for scheduling real-time (RT) tasks includes receiving a task request defined by at least one RT task, wherein the at least one real-time (RT) task is to be queued based on an entry event trigger; obtaining a completion time for the at least one RT task based on an execution deadline of the at least one RT task, an execution deadline of a next RT task, and a maximum execution time of execution of the next RT task; creating a bandwidth reservation task having a deadline defined by the completion time; inserting the at least one real-time task with the bandwidth reservation task into an RT wait queue based on a deadline of each of the at least one RT task and the bandwidth reservation task according to an early deadline first (EDF) criterion; and scheduling unscheduled tasks based on an available time of the RT wait queue according to an EDF-based scheduling.

Disclosed herein are methods and systems for E2E service decomposition to orchestrate and provision services across multiple network domains (a plurality of network domains), and time-aware calendarization. The methods and systems provide time-aware E2E service reservation (e.g., bandwidth allocation) and provisioning across the multiple network domains based on a method and system model that manages the multiple network domains. and methods for verifying and reserving bandwidth and other E2E service reservations.

The application relates to a multi-path fixed time slot and bandwidth reservation scheduling method based on a high-performance network. According to the reservation bandwidth condition in the network resource, the scheduling request sequence is dynamically adjusted to specify a more reasonable scheduling sequence. On the basis of the single-path reservation scheme, the double-path reservation scheme is supplemented, the number of successful request reservations is increased, the requirement for the network bandwidth is reduced, and the reservation bandwidth resource rate is increased.

The application provides a bandwidth reservation method, system and network device of an AVB stream. The application monitors the traffic change of a first AVB stream in a statistical window by an AVB stream sending end, so as to dynamically adjust the current reserved bandwidth of each AVB switching device for the first AVB stream by means of the traffic of the first AVB stream in the continuous N statistical windows, the minimum reserved bandwidth MinBW and the maximum reserved bandwidth MaxBW allowed by the first AVB stream, when the traffic change of the first AVB stream in the continuous N statistical windows meets the reserved bandwidth adjustment requirement, thereby realizing dynamic adjustment of the reserved bandwidth of the AVB stream and improving the utilization rate of the reserved bandwidth of the AVB stream.