60 results about "Programmable networking" patented technology

An adaptive adjustment method of packet-level transmission priority for delay guarantee in programmable networks. The method realizes the adaptive adjustment of packet-level transmission priority in programmable networks from three aspects to satisfy the preset transmission delay requirements, namely: adaptive adjustment of priority in programmable switches based on processing delay state and switch queuing state, adaptive adjustment of priority among programmable switches based on transmission delay in upstream switches, and adaptive adjustment of global parameters based on packet transmission delay satisfaction, so as to realize the packet-level transmission delay guarantee in different positions with different time scales. In addition, for a scene with a large number of time-sensitive flows, based on the optimization model and heuristic algorithm, the optimization deployment of time-sensitive flows can be realized in polynomial time under the condition of limited network resources to improve the overall extent of delay satisfaction.

The embodiment of the invention provides a communication mode self-adaptive switching method, device and equipment based on an SDN (Software Defined Network), which are applied to the technical field of communication, and the method comprises the following steps: receiving a real-time monitoring result reported by a channel state monitoring unit, analyzing the monitoring result, and obtaining a current communication demand and channel state information; according to the current communication requirement and the channel state information, whether communication mode switching or transmission resource redistribution needs to be carried out is judged; if communication mode switching or transmission resource redistribution needs to be carried out, a corresponding instruction is generated and sent to the programmable network equipment; therefore, after the programmable network device receives the instruction, corresponding operation is executed, and communication mode switching or transmission resource redistribution is completed. In this way, the limitation of a single communication mode and fixed transmission resource allocation can be effectively overcome, and a more reliable, efficient and flexible communication solution is provided for the modern communication field.

The invention discloses a collaborative resource scheduling method and system based on a programmable network. The method comprises the following steps: performing performance modeling on heterogeneous computing resource data in all computing resource data to obtain a uniform resource performance index; defining target computing resource data required by multi-dimensional resource demand services, and determining matching relationship information between each multi-dimensional resource demand service and the current computer cluster; in the deployment process, according to all the matching relation information, the correspondingly selected multi-dimensional resource demand service is deployed to each computer cluster; determining a shortest deployment path between the routing nodes in each computer cluster based on a Floyd algorithm; and in the actual scheduling process, executing the computing power arrangement scheme generated in the deployment process. According to the collaborative resource scheduling method and system based on the programmable network provided by the embodiment of the invention, the utilization rate of computing power resources in a wide area network can be improved, and a computing power scheduling strategy is dynamically generated.

Methods, apparatus, and systems for adaptive collaborative memory with the assistance of programmable networking devices. Under one example, the programmable networking device is a switch that is deployed in a system or cluster of servers comprising a plurality of nodes. The switch selects one or more nodes to be remote memory server nodes and allocate one or more portions of memory on those nodes to be used as remote memory for one or more remote memory client nodes. The switch receives memory access request messages originating from remote memory client nodes containing indicia identifying memory to be accessed, determines which remote memory server node is to be used for servicing a given memory access request, and sends a memory access request message containing indicia identifying memory to be accessed to the remote memory server node that is determined. The switch also facilitates return of messages containing remote memory access responses to the client nodes.

An apparatus includes a host interface; a network interface; and a programmable circuitry communicably coupled to the host interface and the network interface, the programmable circuitry comprising one or more processors to implement network interface functionality and to: determine portions of a set of computer vision (CV) processes to be deployed on the programmable circuitry and a host device, wherein the host device to be communicably coupled to the programmable network interface device; access instructions to cause the portions of the set of the CV processes to be deployed on the host device and the programmable network interface device; and wherein a media processing portion of the set of the CV processes is to be deployed to the programmable circuitry, and wherein the programmable circuitry is to utilize media processing hardware circuitry hosted by the apparatus to perform the media processing portion.

Techniques described herein address the above challenges that arise when using host executed software to manage vector databases by providing a vector database accelerator and shard management offload logic that is implemented within hardware and by software executed on device processors and programmable data planes of a programmable network interface device. In one embodiment, a programmable network interface device includes infrastructure management circuitry configured to facilitate data access for a neural network inference engine having a distributed data model via dynamic management of a node associated with the neural network inference engine, the node including a database shard of a vector database.

A system and method for performing rolling updates of distributed applications in a software-defined network environment. An application manager coordinates deployment of updated versions of fxDeviceApp and fxCloudApp components to isolated execution environments on distinct network elements, such as edge devices and cloud-based containers or virtual machines. During the update process, the system establishes secure communication tunnels between the updated components using a virtual messaging fabric, verifies the integrity of the updated components before activation, and maintains service continuity by activating the new versions only after successful validation. The original application instances are decommissioned after the updated components are fully operational. The system supports dynamic scaling of execution environments based on load conditions, policy-governed update concurrency and ordering, coordinated deployment across multiple zones, automatic rollback on failure, and audit logging of deployment events. These mechanisms enable secure, resilient, and minimally disruptive application lifecycle management in distributed and programmable networking environments.

The invention discloses a virtual-real combined sky-ground multi-domain heterogeneous network cooperative computing simulation system and a virtual-real combined sky-ground multi-domain heterogeneous network cooperative computing simulation method. The system comprises a heterogeneous computing power prototype subsystem, a virtual simulation subsystem, a satellite-borne general computing terminal, a programmable network simulator, a heterogeneous resource management subsystem and a simulation verification system interface. Real hardware computing power characteristics are introduced through a physical board card, large-scale expansion is realized through virtual nodes, and authenticity and large-scale performance of simulation are considered; link parameters are generated according to a high-precision orbit dynamics model, and the fidelity of sky-ground network simulation is greatly improved; based on a mixed container arrangement technology, unified management and intelligent scheduling of sky, air and ground heterogeneous resources are realized, and a powerful platform is provided for verifying an advanced cross-domain collaborative algorithm; one set of system can support various simulation verification requirements of networking routing, task unloading, distributed cooperative computing, constellation cluster management and control and the like, and the research and development cost and complexity are reduced.

A spiking neural network for classifying input signals is disclosed. The spiking neural network includes a plurality of spiking neurons and a plurality of synaptic elements that interconnect them to form a network. Each synaptic element receives a synaptic input signal and applies a weight to generate a synaptic output signal, and the synaptic element is configurable to adjust the weight applied by each synaptic element. Each spiking neuron is adapted to receive the synaptic output signal from the synaptic element and generate a spatio-temporal spike train output signal in response thereto. The spiking neural network is divided into a plurality of sub-networks, which include a subset of spiking neurons connected to receive the synaptic output signal from a subset of synaptic elements. The sub-network is adapted to generate a sub-network output pattern signal in response to a sub-network input pattern signal. In addition, each sub-network forms part of one or more cores in a core array, and each core consists of a programmable network of spiking neurons implemented in hardware or a combination of hardware and software. Communication between cores in the core array is arranged through a programmable interconnect structure.

An apparatus to facilitate metrics and security-based accelerator service rescheduling and auto-scaling using a programmable network device is disclosed. The apparatus includes processors to collect metrics corresponding to communication links between microservices of a service managed by a service mesh; determine, based on analysis of the metrics, that a workload of the service can be accelerated by offload to a hardware accelerator device; generate a scaling request to cause the hardware accelerator device to be allocated to a cluster of hardware devices configured for the service; cause the scaling request to be transmitted to a programmable network device managing the hardware accelerator device, the programmable network device to allocate the hardware accelerator device to the cluster and to register the hardware accelerator device with the service mesh; and schedule the workload of the service to the hardware accelerator device.

The invention discloses an anomaly detection positioning method and system based on a programmable network, and a medium, and belongs to the technical field of data processing. The method comprises the steps that S1, INT header information is added to an entry node, a first data packet passes through an INT source switch, an INT header label is added to the first data packet, a metadata field is added, and a second data packet is obtained; s2, if the intermediate node is abnormal, adding abnormal information to the second data packet to obtain a third data packet, and outputting the third data packet to a next-hop switch; s3, when the final data packet arrives at the last hop of switch, judging whether the final data packet needs to be sent to the INT server by using the Bloom filter, and finally updating the Bloom filter; and S4, after each detection period is finished, clearing the Bloom filter and the hash table of the switch. Abnormality judgment and recording are carried out on the data plane, so that the method can directly carry out anomaly detection in the transmission process of the data stream, and the data volume sent to the INT server is remarkably reduced.

Various example embodiments of a programmable network slice framework may be configured to support programmable network slices in a radio access network (RAN) portion of a programmable RAN slice framework-based cellular communication system. Various embodiments of a programmable RAN slice framework may be configured to support a programmable network slice in a RAN based on using a set of RAN slices having related network slice operating states and related network slice configurations that may be used to allocate RAN resources to user equipment (UE) of the RAN, frequency domain resources of the RAN slice are allocated to UEs of the RAN based on a network slice operating state and a network slice configuration of the RAN slice based on the use of frequency domain scheduling capabilities, various aspects of the RAN slice are supported based on the use of various types of messaging, and the like.

A system for performing learning is described. The system includes a linear programmable network layer and a nonlinear activation layer. The linear programmable network layer includes an input, an output, and a linear programmable network component interconnected between the input and the output. The nonlinear activation layer is coupled to the output. The linear programmable network layer and the nonlinear activation layer are configured to have a static state at a minimum value of the content of the system.

The invention relates to an in-network neural network deployment method based on an SDN modal, and the method comprises the steps: mapping an input feature and an output result of a deep neural network model into a flow table item of network equipment, achieving the deep neural network reasoning through a table item matching mode, and achieving the network functions of flow classification, anomaly detection and the like. According to the method, the deep neural network model is compressed and converted into the table item matching task suitable for the programmable network equipment, so that efficient network data plane reasoning is realized. According to the method, the utilization of storage and computing resources is optimized, the computing complexity of network equipment is remarkably reduced, and the throughput and the real-time performance are improved. Through an innovative table item compression and adaptive matching mechanism, the method supports high-precision applications such as traffic classification and anomaly detection, and the network intelligent management and optimization capability is remarkably improved. According to the method, efficient deployment of the deep neural network in the in-network equipment can be realized, and an effective solution is provided for development of an intelligent network technology.

Embodiments of the invention provide a computer-implemented method that includes intercepting, using a controller of a programmable network (PN), a deployment of a control application in the PN. Responsive to intercepting the deployment of the control application, a user experience (UE) analysis is performed. The UE analysis includes determining, based at least in part on a set of digital twins, a UE impact of the control application; and, based at least in part on the UE impact, generating a UE-based control application deployment recommendation. The UE analysis further includes deploying the control application based at least in part on the UE-based control application deployment recommendation.

An apparatus includes a host interface, a network interface, and a programmable circuitry communicably coupled to the host interface and the network interface. The programmable circuitry can include one or more processors to implement network interface functionality, and a discrete trusted platform module (dTPM) to enable the one or more processors to establish a secure boot mechanism for the apparatus, wherein the one or more processors are to instantiate a virtual TPM (vTPM) manager that is associated with the dTPM, the vTPM manager to host vTPM instances corresponding to one or more virtualized environments hosted on at least one of the programmable circuitry or a host device communicable coupled to the apparatus.

The application provides an online dynamic programming method, system and storage medium of a programmable network device, comprising: a control terminal calling a network service function definition to generate a network service instruction sequence; based on the network service instruction sequence and a self-defined protocol structure, a self-defined protocol header is constructed and inserted into an IP data packet, and a type field of the IP data packet is marked with a mark value that the current data packet belongs to a normal data packet or contains network service instructions, and the IP data packet is sent to the programmable network device; wherein the self-defined protocol structure comprises a protocol type, a number of instructions and an instruction sequence; the programmable network device receives the IP data packet, when the mark value in the IP data packet indicates that the current data packet contains network service instructions, the self-defined protocol header is parsed, the network service instruction sequence is executed and removed one by one, and the network service of the programmable network device is dynamically adjusted, otherwise the IP data packet is directly forwarded. The application can solve the problem that the existing programmable network technology cannot simultaneously meet high performance and online programming flexibility.

A system and method for performing load balancing in an In-Network Aggregation (INA) application is provided. A network server partitions data to be aggregated into data parts and sends the data parts in first packets to programmable network switches. Each programmable network switch receives a plurality of first packets, aggregates the first data parts to generate an aggregated data part, and sends the aggregated data part in a second packet to an aggregation network server. The aggregation network server receives a plurality of second packets and generates aggregated data from the aggregated data parts. Routing headers of the first packets may include the programmable network switch address as an intermediate waypoint and the aggregation network server address as a destination address. The routing headers may include a second programmable network switch address as a second intermediate waypoint. The data to be aggregated may be value vectors or key-value pairs.

The disclosure features a method of creating a communications network. In this method, a first tape node includes a first type of wireless communication interface and a second type of wireless communication interface having a longer range than the first type of wireless communication interface. A second tape node includes the first type of wireless communication interface that is operable to communicate with the first tape node over a wireless communication connection established between the first type of wireless communication interfaces of the first and second tape nodes. Over a wireless communication connection, the first tape node sends programmatic code executable by the first tape node to function as a master tape node with respect to the second tape node.

A method is implemented by a programmable network device, where the method is to enable debugging of a parser in a packet processing pipeline of the programmable network device. The method includes receiving a packet for processing in the packet processing pipeline, initializing a tracing structure to record debugging information for the parser, adding a current state of the parser to the tracing structure, and determining a next state of the parser.

The present invention discloses a method and system for deploying backup paths for programmable network fault recovery. The method comprises: obtaining the switch centrality using betweenness centrality based on the latency and reliability requirements of nodes in the network; calculating the network resource consumption of the backup path based on latency, reliability, and bandwidth to obtain the backup path weight; obtaining the optimal backup path for important switch nodes in the network using a heuristic algorithm based on network resources and user requirements for the backup path; and the control plane distributing the optimal backup path set to the programmable computing plane to complete the deployment of the backup path. The present invention can provide the network with backup paths that meet latency and reliability requirements and are as few as possible within polynomial time, thereby improving network storage resource utilization while meeting requirements such as low latency, high available load, and high reliability. The present invention has a short time to obtain the backup path, is simple to deploy, is flexible to implement, and has strong practicality.

The application discloses a large-scale traffic generation method and system based on programmable network technology, which is used for the research on network operation and maintenance and the defense of DDoS attack and the like. The method generates the required large-scale traffic on demand through a coordination server and a programmable network switch. A server group is used to generate data packets of the large-scale traffic and deliver the data packets to the programmable switch, and the programmable switch controls the sending speed of the large-scale traffic, so that the low-cost and high-extensibility on-demand generation of the large-scale traffic is realized. Specifically, a series of primitives based on intention and irrelevant to underlying architecture details are designed, so as to reduce the difficulty of describing the intention of generating the large-scale traffic; and a coordination mechanism of the server and the programmable switch is designed, so that the required configuration can be completed on the switch and the server according to the intention expressed by different kinds of primitives, and the generation of the large-scale traffic is realized by coordinating and utilizing the server and the switch resources.

The present invention discloses a method and interface for extracting traffic features based on a programmable network, relating to the technical field of traffic feature extraction. The specific steps are: abstracting data packets to obtain data packet tuples; grouping the data packet tuples according to a specified granularity to obtain group tuples; filtering the data packet tuples or the group tuples based on preset filtering conditions to obtain filtered tuples; mapping the filtered tuples to generate mapping tuples, performing aggregation calculations on the mapping tuples to obtain feature sequences, and converting the feature sequences into feature vectors. The present invention can simplify the underlying hardware implementation details and reduce complexity, while closely integrating with the application characteristics of traffic analysis to provide a universal feature extraction method. This method is of great significance for improving the overall performance of network traffic analysis systems and promoting the development of network security technology.

The application discloses an abnormality detection positioning method and system based on a programmable network and a medium, and belongs to the technical field of data processing. The method comprises the following steps: S1, adding INT header information at an entry node, and passing a first data packet through an INT source switch to obtain a second data packet by adding an INT header label and a metadata field to the first data packet; S2, if an intermediate node is abnormal, adding abnormality information to the second data packet to obtain a third data packet and outputting the third data packet to a next-hop switch; S3, when a final data packet reaches a last-hop switch, using a Bloom filter to determine whether the final data packet needs to be sent to an INT server, and finally updating the Bloom filter; and S4, emptying the Bloom filter and a hash table of the switch after each detection period ends. Abnormality judgment and recording are performed on a data plane, so that the application can directly perform abnormality detection in the transmission process of a data stream, and thus the amount of data sent to the INT server is significantly reduced.

The invention discloses a program optimization method based on an LLVM framework and SMT / ILP solution, which realizes efficient program synthesis under hardware resource constraint through two stages of intermediate code generation and compilation optimization. The front end divides transaction statement blocks and optimizes dependence, and the rear end generates P4 codes adaptive to the target platform through mathematical modeling and solving, so that resource occupation and debugging cost are remarkably reduced, and the method is suitable for program development of programmable network equipment.

This invention discloses a distributed data plane resource optimization method and system, which fully utilizes merging and splitting algorithms to optimize data plane program resources and solve the resource limitation problem of switches in programmable networks. First, the input programs are classified. Then, a merging algorithm is used for resource-sparse programs, and a splitting algorithm is used for resource-intensive programs. The merging and splitting algorithms yield a set of subtasks, and finally, the deployment scheme of each task in the subtask set in the underlying network is obtained. Compared with existing technologies, this invention can improve the speed and effectiveness of program merging, effectively reduce program redundancy, and reduce the granularity of program placement and scheduling overhead between devices by splitting programs. It also provides a deployment scheme with higher resource utilization, demonstrating the practicality of a distributed data plane resource optimization method and system.

The invention relates to the technical field of programmable networks, in particular to an extensible programmable network verification method and system, and the method comprises the steps: describing a network attribute in a network program needing to be verified through an assertion statement, and adding a corresponding annotation in the network program needing to be verified; a symbolic execution optimization strategy is used for optimizing the annotated network program needing to be verified, and a C language translator is used for modeling the optimized network program needing to be verified into an equivalent C language; and verifying the equivalent C language by using a C language symbolic execution engine to obtain an assertion execution result used for describing whether the network program to be verified is correct or not. According to the method, the assertion language specification is provided for the user by utilizing the assertion statement, and the complexity of network expected expression is reduced; the symbolic execution optimization strategy is used for simplifying the strategy selection process of developers, the correctness of network operation logic is ensured, a large-scale network can be verified in a short time, and the method has important significance for improving the safety, reliability and performance of the programmable network.

Programmable network devices configured to perform various packet processing functions are further configured to use IPsec to secure control and data packets associated with data sessions traversing the UPF. Field-programmable gate arrays (FPGAs) and / or graphics processing units (GPUs) coupled with network interfaces are configured to perform user plane functions and secure data packet within a compact and modular hardware unit to minimize excessive communication while maintaining control and user plane separation (CUPS).