A deterministic network service method based on business characteristic perception and related device

Abstract

The present application discloses a deterministic network service method and related devices based on business characteristic perception, which relates to the field of network service technology. The method includes obtaining the global state information of the network; the global state information includes: packet loss rate, delay, jitter and available bandwidth; when a new data packet needs to be transmitted in the network, the business type and corresponding business demand of the new data packet are determined; according to the business type and corresponding business demand of the new data packet, resources are dynamically matched; when the current network has sufficient capacity to support the business demand of the new data packet, the optimal transmission path is calculated according to the global state information of the network; according to the business type and the optimal transmission path of the new data packet, the new data packet is allocated to the corresponding forwarding queue. The present application can meet the personalized needs of specific businesses such as the industrial Internet for low network latency, high reliability, wide coverage and large data rate.

Description

Technical Field

[0001] The present application relates to the field of network service technology, and in particular to a deterministic network service method and related devices based on business characteristic perception. Background Art

[0002] In recent years, emerging businesses such as smart transportation, environmental monitoring, and power grid inspection have developed rapidly, and the demand for networks has continued to grow. Traditional ground networks are difficult to meet the needs of wide-area coverage and high performance. Emerging business scenarios have put forward higher requirements for networks, including higher data transmission efficiency, lower latency, and wider coverage. In addition, different industry applications have different requirements for networks. For example, power networks require highly reliable and low-latency networks to meet the power system's requirements for real-time and stability. In contrast, remote ships require a wide-coverage network to ensure reliable and continuous network support under marine environmental conditions.

[0003] At present, the networks providing services are mainly composed of ground cellular communication systems and satellite communication systems. These systems have developed independently due to technical and market reasons, and it is difficult to achieve coordinated guarantee of service quality of business networks. There are two main reasons for the above: First, it is difficult to obtain the real-time status of the network. Although different manufacturers have proposed different network monitoring protocols, the protocols are difficult to be compatible, which increases the difficulty of perceiving the overall network performance. At the same time, traditional routing protocols usually use the shortest path as the main selection criterion, which performs well in ground fixed networks, but is difficult to apply to heterogeneous networks in the sky and ground. These protocols usually perform routing selection based on static network status information, cannot adapt to the dynamically changing network environment in a timely manner, lack flexibility, and are difficult to meet the diverse requirements of different applications for quality of service (QoS). Therefore, network design and optimization must be combined with specific industry applications to meet their unique needs.

[0004] Current network awareness implementations mainly rely on network monitoring solutions and protocols provided by network equipment manufacturers. These solutions and protocols are targeted at specific devices and can obtain network performance data from different dimensions. For example, sFlow, created by InMon, is a network monitoring technology based on traffic sampling that can monitor a large amount of traffic in high-performance networks in real time. It provides a comprehensive view of network traffic by collecting samples of data packets on switches and routers. The advantages of sFlow are its high efficiency and low overhead. It is suitable for monitoring large-scale, high-speed network environments and is often used for traffic analysis, security monitoring, and network performance optimization. NetFlow, developed by Cisco, is a traffic monitoring technology used to record traffic information in the network in detail. NetFlow captures and analyzes flow records to provide detailed information about traffic sources, destinations, protocols, and applications. It is widely used in network traffic monitoring, performance analysis, bandwidth planning, and security detection. NetFlow's detailed data recording capabilities make it the basis for many network management and analysis tools. NetStream, developed by Huawei, is a traffic monitoring technology similar to NetFlow that collects and analyzes network traffic data and provides detailed information about traffic patterns and behaviors. Huawei's NetStream supports sophisticated traffic monitoring and analysis, which helps optimize network performance, traffic engineering, and security management. It is similar to NetFlow in terms of functions and applications, but is more in line with the characteristics of Huawei's equipment and solutions.

[0005] In order to meet the personalized needs of the business for the network, a lot of research has been done in homogeneous networks. In order to optimize the transmission efficiency and reliability of the power communication network, Liu Sifang proposed a credit shaping transmission mechanism based on business priority. This mechanism ensures the transmission priority of high-priority services by allocating credits, while reasonably scheduling low-priority services. Specifically, the credit represents the transmission weight of the service in the network. High-priority services obtain more credits and can still be guaranteed when resources are tight. The mechanism introduces traffic shaping technology to avoid network overload by controlling the traffic rate. During the transmission process, network nodes manage queues according to the credits, give priority to high-credit services, and ensure the timeliness and reliability of their transmission. In the satellite network, Du Jianan divided the three services into three QoS demand types, namely high latency demand (Class A), high bandwidth demand (Class B), and low latency bandwidth demand (Class C), and proposed an inter-satellite routing algorithm based on service scheduling. Class A and B have high priority and adopt a low-complexity routing strategy with priority forwarding within the orbit. Class C services adopt an inter-satellite routing strategy based on link status. After combining the two, the congestion rate and packet loss rate increase with the number of service flows more slowly than the common algorithms.

[0006] At present, some research has also been conducted on the coordinated use of satellite-ground network resources for business transmission optimization. Ma proposed a low-latency distributed collaborative computing strategy based on computation-dependent routing for satellite-ground integrated network (STIN). This strategy uses an algorithm called directed diffusion enhanced task scheduling to achieve distributed computing and transmission while considering computing and communication resources. Based on the dynamic characteristics of STIN and the complex changes in link status, Liu Jinlin proposed using a deep reinforcement learning method to dynamically adjust the routing strategy to optimize network performance through interactive learning between the agent and the environment. In the algorithm design, the agent observes the state space containing the link quality and uses the next hop candidate transmission node as the action space, gradually learning a routing strategy that can adaptively perceive the link status and dynamically adjust. The agent uses historical data to learn the rules and adjusts the strategy according to the feedback from the environment, thereby optimizing the end-to-end delay and avoiding interference and congestion.

[0007] However, the prior art has the following disadvantages.

[0008] 1. Difficulty in real-time network status monitoring: Different management agencies and equipment manufacturers in traditional networks have proposed their own network performance monitoring technologies, but the various solutions are difficult to be compatible. In addition, the network perception of the above solutions is poor in real time, and it is difficult to ensure the monitoring accuracy, the effect and quality of obtaining network status are difficult to guarantee.

[0009] 2. Inefficient dynamic resource management: The complex and ever-changing network application scenarios such as the Industrial Internet require dynamic scheduling and efficient utilization of network resources. Existing technologies have not yet reached an ideal level in terms of real-time perception of network status, changes in user demand, and dynamic scheduling of resources.

[0010] 3. Insufficient support for personalized network services: The communication equipment of the Industrial Internet is diverse, and the demand for personalized network services is high. Existing network technologies still face many challenges in providing customized solutions for different industrial applications.

[0011] Therefore, how to design a network service mechanism that can meet the personalized needs of specific businesses such as the industrial Internet for low network latency, high reliability, wide coverage, and large data rates has become a technical problem that needs to be urgently solved in this field. Summary of the invention

[0012] The purpose of this application is to provide a deterministic network service method and related devices based on business characteristic perception, which can meet the personalized needs of specific businesses such as the industrial Internet for low network latency, high reliability, wide coverage and large data rate.

[0013] To achieve the above objectives, this application provides the following solutions.

[0014] In a first aspect, the present application provides a deterministic network service method based on business characteristic perception, characterized in that the deterministic network service method based on business characteristic perception includes.

[0015] Obtain global status information of the network; the global status information includes: packet loss rate, delay, jitter and available bandwidth.

[0016] When a new data packet needs to be transmitted in the network, the service type and corresponding service requirements of the new data packet are determined.

[0017] Dynamic resource matching is performed according to the service type of the new data packet and the corresponding service demand.

[0018] When the current network has sufficient capacity to support the service demand of the new data packet, the optimal transmission path is calculated based on the global state information of the network.

[0019] According to the service type of the new data packet and the optimal transmission path, the new data packet is allocated to a corresponding forwarding queue.

[0020] In a second aspect, the present application provides a deterministic network service device based on business characteristic perception, characterized in that the deterministic network service device based on business characteristic perception includes.

[0021] The information acquisition module is used to acquire the global status information of the network; the global status information includes: packet loss rate, delay, jitter and available bandwidth.

[0022] The service type and service requirement determination module is used to determine the service type and corresponding service requirement of a new data packet when a new data packet needs to be transmitted in the network.

[0023] The resource dynamic matching module is used to dynamically match resources according to the service type of the new data packet and the corresponding service demand.

[0024] The optimal transmission path determination module is used to calculate the optimal transmission path according to the global state information of the network when the current network has sufficient capacity to support the business needs of new data packets.

[0025] The allocation module is used to allocate the new data packet to a corresponding forwarding queue according to the service type of the new data packet and the optimal transmission path.

[0026] In a third aspect, the present application provides a computer device, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement any of the above-mentioned deterministic network service methods based on business characteristic perception.

[0027] In a fourth aspect, the present application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements any of the above-mentioned deterministic network service methods based on business characteristic perception.

[0028] In a fifth aspect, the present application provides a computer program product, including a computer program, which, when executed by a processor, implements any of the above-mentioned deterministic network service methods based on business characteristic perception.

[0029] According to the specific embodiments provided in this application, this application discloses the following technical effects:

[0030] The present application provides a deterministic network service method and related device based on business characteristic perception. First, the global state information of the network is obtained, and the global state information includes: packet loss rate, delay, jitter and available bandwidth; the present application actively sends a detection data packet to the network to ensure that the detection data packet is the same as the business flow path, and flexibly analyzes the network behavior according to the characteristic changes of the detection data packet affected by the integrated network. Then when a new data packet needs to be transmitted in the network, the business type and corresponding business requirements of the new data packet are determined; according to the business type and corresponding business requirements of the new data packet, dynamic resource matching is performed; the task demand-driven scheduling mechanism identifies the task importance index and combines node resources and business requests for scheduling, thereby meeting the industry's network requirements for specific tasks and improving the differentiated service level of the business. Finally, when the current network has sufficient capacity to support the business requirements of the new data packet, the optimal transmission path is calculated according to the global state information of the network; according to the business type of the new data packet and the optimal transmission path, the new data packet is assigned to the corresponding forwarding queue. This application is based on the perception of industry network needs and signal measurement information such as network latency, and ensures the end-to-end latency and other performance indicators of each service through mechanisms such as deterministic routing, thereby meeting the deterministic transmission quality requirements of specific network transmission services such as industrial control. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0032] Figure 1 A flowchart of a deterministic network service method based on business characteristic perception is provided in one embodiment of the present application.

[0033] Figure 2 The present invention provides a schematic diagram of the overall process of a deterministic network service method based on business characteristic perception according to an embodiment of the present application.

[0034] Figure 3 A schematic diagram of a network status collection method provided in an embodiment of the present application.

[0035] Figure 4 A collection flow chart provided for an embodiment of the present application.

[0036] Figure 5 A schematic diagram of an existing network compatible packet header provided in an embodiment of the present application.

[0037] Figure 6 A schematic diagram of network status perception provided by an embodiment of the present application.

[0038] Figure 7 A schematic diagram of a scheduling mechanism provided for an embodiment of the present application.

[0039] Figure 8 A schematic diagram of a business execution process provided in an embodiment of the present application.

[0040] Fig. 9 A schematic diagram of a multi-priority queue structure provided in one embodiment of the present application.

[0041] Fig.10 A schematic diagram of a queue scheduling process provided in one embodiment of the present application.

[0042] Fig.11 A schematic diagram of a route optimization process provided for an embodiment of the present application.

[0043] Fig.12 A schematic diagram of functional modules of a deterministic network service device based on business characteristic perception provided by an embodiment of the present application.

[0044] Fig.13 A schematic diagram of the structure of a computer device provided in one embodiment of the present application. DETAILED DESCRIPTION

[0045] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0046] In order to make the above-mentioned objects, features and advantages of the present application more obvious and easy to understand, the present application is further described in detail below in conjunction with the accompanying drawings and specific implementation methods.

[0047] Based on the above background, this application proposes a deterministic network service mechanism based on business characteristic perception, which aims to meet the personalized needs of specific businesses such as the industrial Internet for low network latency, high reliability, wide coverage, and large data rates. Through real-time perception of multi-dimensional resources of the integrated network, unified resource collaborative management and adaptation, and high-reliability adaptive routing technology, it is possible to provide network transmission services that meet the needs of specific demand businesses.

[0048] The definitions of abbreviations and key terms are shown in Table 1.

[0049] Table 1 Abbreviations and definitions of key terms

[0050]

[0051] In an exemplary embodiment, Figure 1 As shown, a deterministic network service method based on business characteristic perception is provided, and the method includes the following steps S1 to S5.

[0052] S1: Obtain global status information of the network; the global status information includes: packet loss rate, delay, jitter and available bandwidth.

[0053] S2: When a new data packet needs to be transmitted in the network, the service type and corresponding service requirements of the new data packet are determined.

[0054] S3: Dynamically matching resources according to the service type of the new data packet and the corresponding service demand.

[0055] S4: When the current network has sufficient capacity to support the service demand of the new data packet, the optimal transmission path is calculated according to the global state information of the network.

[0056] S5: Allocate the new data packet to a corresponding forwarding queue according to the service type of the new data packet and the optimal transmission path.

[0057] Systems such as autonomous driving and remote control are being widely deployed and applied around the world. The operation of these systems is extremely dependent on stable and efficient network services. Therefore, network optimization based on business needs is crucial to ensure the real-time and stability of such systems. This application proposes a new method based on the combination of network perception and business perception to optimize the quality of network services, such as Figure 2 As shown in the figure, deep network perception is used to perceive the state changes in the network in real time, precise business perception is used to identify the type of business flow carried, and dynamic resource scheduling and routing optimization mechanisms ensure efficient network services at different levels.

[0058] Throughout the process, the deep network perception module constructs a detection data packet and forwards it across the entire network to obtain the global state information of the network. Subsequently, the global state information is processed and stored in the database. When a new data packet needs to be transmitted in the network, the business precision perception module parses the data packet through the port, identifies the business type it carries, and determines its specific needs based on the business type. When the data packet transmission starts, the resource dynamic scheduling module first performs dynamic matching of resources. If the current network has sufficient capacity to support the transmission needs of the business, the routing optimization mechanism will use the real-time network status data provided by the deep network perception module to calculate the end-to-end transmission path that meets the business needs from the entire network, and finally select the optimal path for data transmission. When the data packet is forwarded through the switch along the specified path, the switch will assign the data packet to the corresponding forwarding queue according to the business type to ensure priority processing and efficient transmission.

[0059] 1. Deep network perception.

[0060] Since different networks have different monitoring methods for network resources, in order to simplify the complexity of perception, the network is perceived in an active measurement manner. Specifically, by actively sending detection packets to the network, the network behavior is analyzed based on the characteristic changes of the detection packets affected by the network. Figure 3 As shown in the figure, the acquisition flow chart is as follows Figure 4 shown.

[0061] Step 1: When a normal data message reaches the first switching node of the system, the status acquisition module matches and mirrors the message through the sampling method set on the switch, inserts a detection header before the payload field of the message according to the needs of the telemetry task, and encapsulates the telemetry information specified by the detection header into metadata MD and inserts it after the detection header.

[0062] Step 2: When the message is forwarded to the intermediate node, the device matches the detection header and inserts the MD.

[0063] Step 3: When the message is forwarded to the last hop of the in-band network telemetry system, the switching device matches the detection header and inserts the MD of the local device. After that, all MDs in the detection data packet are extracted and encapsulated with an IP header to form a telemetry data packet. The outer IP is the telemetry server address, so the telemetry data packet is forwarded to the telemetry server. The telemetry server parses the telemetry information in the telemetry data packet and reports it to the upper-layer telemetry application.

[0064] The processing of detection data packets by network nodes needs to be realized through the programming of data packet headers, but some devices that have been put into use do not support programming. A universal data packet format is needed to be compatible with traditional existing network devices. Compatible packet headers such as Figure 5 shown.

[0065] When there are nodes in the network where the detection packet passes that do not support programmability, the steps for network status perception are as follows: Figure 6 shown.

[0066] Step 1: The first switching node reached by the normal data message.

[0067] Step 2: Determine whether the node supports programmability.

[0068] Step 3: If programmable is supported, use Figure 4 Process according to the process shown in step 1 in.

[0069] Step 4: If not supported, perform traditional IP forwarding and forward using the IP address.

[0070] Step 5: When the message is forwarded to the intermediate node, repeat steps 2, 3, and 4.

[0071] Step 6: When reaching the last hop switch, determine whether the device supports programmability.

[0072] Step 7: If supported, insert the required status information and encapsulate the IP header.

[0073] Step 8: If not supported, directly encapsulate the IP header.

[0074] Step 9: The telemetry server parses the telemetry information in the telemetry message and reports it to the upper-layer telemetry application.

[0075] In an exemplary embodiment, step S1 specifically includes:

[0076] S11: Construct a detection data packet.

[0077] S12: forwarding the detection data packet in the entire network to obtain global status information of the network.

[0078] In step S12, it specifically includes.

[0079] S121: When the detection data packet reaches the first switch, determine whether the first switch supports programmability, and obtain a first determination result.

[0080] S122: If the first judgment result is yes, mirror the detection data packet to obtain a mirrored original message.

[0081] S123: inserting required network information into the mirrored original message and forwarding it to the next-hop switch; the required network information is metadata encapsulated from the telemetry information specified by the detection header.

[0082] S124: When the next-hop switch matches the detection header, the required network information is inserted, and it is determined whether the last-hop switch is reached to obtain a second determination result.

[0083] S125: If the second judgment result is no, return to the step of "forwarding to the next-hop switch".

[0084] S126: If the second judgment result is yes, determine whether the last-hop switch supports programmability to obtain a third judgment result.

[0085] S127: If the third judgment result is yes, the last-hop switch inserts the required network information after matching the detection header, and encapsulates an IP header; the IP is the telemetry server address.

[0086] S128: If the third judgment result is no, directly encapsulate the IP header.

[0087] S129: If the first judgment result is no, traditional IP forwarding is adopted, and the process returns to the step of “forwarding to the next hop switch”.

[0088] S1210: Parse the telemetry information according to the IP header to obtain the global status information of the network, and report it to the upper-layer telemetry application.

[0089] 2. Accurate business perception.

[0090] In view of the current new businesses such as telemedicine, smart factories and personalized streaming media, this application divides the network requirements into three categories as a whole: low latency, large bandwidth and best effort delivery. Taking industrial Internet of Things control as an example, it involves low-latency transmission of industrial control information, large-bandwidth transmission of real-time monitoring data and transmission of industrial Internet management information. Among them, when the remote control center issues control instructions to specific equipment, Modbus TCP on port 502 is used to issue instructions to the programmable logic controller; real-time monitoring data transmission enables the control center administrator to monitor the real-time picture of the site, and establishes an RSTP stream with port 554 of the camera recorder to obtain real-time pictures; industrial Internet management information transmission mainly realizes the control center's access to the status of specific control devices, using the HTTP protocol on port 80, and the corresponding network transmission requirements are shown in Table 2.

[0091] Table 2 Network requirement mapping relationship

[0092]

[0093] By parsing the packet header of P4, the value of the port field in the transport layer can be identified, and the type of service transmitted by the data packet and its network transmission requirements can be understood according to the above mapping relationship.

[0094] In an exemplary embodiment, step S2 specifically includes:

[0095] S21: parse the new data packet according to the port to obtain the service type of the new data packet.

[0096] S22: Determine corresponding specific requirements according to the service type of the new data packet.

[0097] 3. Dynamic resource scheduling.

[0098] The unified resource management and control of the current heterogeneous networks is complex and it is difficult to support the coordinated and efficient adaptation and application of resources for complex businesses. In response to the above problems, this application designs a method to realize the network requirements of different industries for specific services, and designs a scheduling mechanism that can be based on business importance indicators. The overall architecture capability resources are divided into four layers. The first layer is the management layer, which maintains a processing queue, and will temporarily join the queue when a specific business arrives; the second layer is the coordination layer, which contains multiple coordinators, which can coordinate different sub-executors according to the business type, and call different sub-executors according to the sensitivity of different businesses to different resources; the third layer is the planning layer, which schedules resource layer resources according to the scheduling strategy of the coordination layer; the fourth layer is the resource layer, which manages heterogeneous network resources, such as Figure 7 shown.

[0099] The architecture actually builds a producer-consumer model, which decouples the executor and the business. There is no direct connection between them, so different execution strategies can be selected according to the business type. There are three execution strategies: (1) Buffered execution. Buffered execution is used to execute time-insensitive businesses. The task volume in this scenario is huge and does not need to be completed instantly. Instead, it focuses on how to use limited resources to process as many tasks as possible in a unit time, that is, the problem of throughput priority; (2) Immediate execution. This type of business often needs to be executed as quickly as possible, and the higher priority needs to be executed immediately; (3) Rejection execution. Rejection execution is the protection strategy of the architecture. Rejection strategies can also be divided into the following three types according to different businesses: (1) Direct abandonment, (2) Discarding the business at the front of the queue and then resubmitting the rejected business, and (3) The caller executes the business.

[0100] In terms of system resource allocation, business priorities are Expressed as a function of multiple business characteristics, including the urgency of the business , Resource Requirements and expected execution time The priority calculation formula is as follows.

[0101] .

[0102] in, Indicates business The higher the urgency, the faster the business needs to be processed. Positive impact on priorities. Resource requirements Adopted This means that the greater the resource demand of the business, the smaller its priority contribution. This processing method avoids the business with excessive resource consumption occupying the limited resources in the system, thereby ensuring the reasonable allocation of overall resources under limited network resources. In addition, the expected execution time Used The form of indicates that the longer the expected execution time is, the smaller its positive contribution to the priority is, avoiding long-term business blocking the system and ensuring that urgent short-term tasks can be processed first. , and The coefficients represent the weight of each factor, which is used to flexibly adjust the impact of different business attributes on the priority. These coefficients can be adjusted according to specific business scenarios to ensure the flexibility and adaptability of priority calculation.

[0103] Calculating business priorities After that, the next step is to allocate resources according to priority and resource demand. This application proposes a resource allocation function , which is used to describe the coordinator How to prioritize your business and actual resource requirements For the child executor Allocate resources.

[0104] .

[0105] In this formula, the numerator Indicates business The total actual resource demand adjusted according to its priority. The higher the priority of the business, the more resources it needs, allowing it to obtain more resources during the allocation process. This ensures that high-priority businesses can be met in a timely manner and meet the system's response requirements for emergency businesses. Representation Coordinator The sum of resource requirements of all managed child executors, where Is a coordinator The entire business portfolio undertaken, and for This design ensures that in the overall resource pool, each business is allocated resources according to demand, ensuring fairness in resource allocation and efficient use of system resources.

[0106] In the above resource allocation model, by calculating the priority of each business Combined with actual resource requirements The dynamic allocation of resources is realized. The efficiency of resource allocation is the key to the performance of the entire system. Resource utilization is an important indicator to measure this efficiency. It can be defined by the ratio of the actual amount of resources used to the total amount of available resources. The calculation formula is as follows.

[0107] .

[0108] in, Represents a sub-executor The actual amount of resources used, It is the total amount of resources in the system. Through resource utilization, we can evaluate the rationality of resource allocation in the system and ensure the efficiency of resource allocation among different businesses.

[0109] To avoid excessive resource usage and system performance degradation, set queue length limits and service rejection policies. Set thresholds , when the queue length achieve When adopting a rejection strategy , to prevent excessive backlog of tasks. When business execution requires waiting, the business execution waiting time needs to be calculated , which can be estimated by the number of tasks in the queue and the processing capacity of the system. The waiting time can be calculated cumulatively based on the processing time of all tasks before it.

[0110] .

[0111] In the above formula, For Business A business that needs to be processed before, Indicates business processing time, The priority of the business. Processing time It depends on the amount of resources allocated by the system to the business , can be achieved through the function Dynamic adjustment.

[0112] .

[0113] This function It reflects the impact of resource allocation on task execution time. The more resources are allocated, the faster the task is processed, thus shortening the time. .

[0114] In terms of algorithm implementation, the overall processing flow is as follows Figure 8 shown.

[0115] Step 1: When a business arrives, first determine whether the execution pool is empty. If it is empty, then determine whether the business queue is full. Otherwise, perform task type classification.

[0116] Step 2: If the service queue is full, a rejection strategy is adopted; otherwise, the service is added to the queue.

[0117] Step 3: The coordinator divides the business types. If the division fails, it throws the process, otherwise it proceeds to the next step.

[0118] Step 4: Distribute specific services to sub-executors and determine whether the corresponding sub-executors are full.

[0119] Step 5: If the sub-executor is full, wait. If the waiting time exceeds the preset time, throw the process.

[0120] Step 6: If the sub-executor is not full, assign the specific business to the sub-executor and release the resources after completion.

[0121] For data packets transmitted over the network, different services are classified by analyzing the fields in the data packet header, thereby achieving accurate perception of the services and helping to clarify the specific network requirements of different services. On this basis, by matching the queue scheduling measurement inside the network node with the data packet transmission requirements of the service requirements, the transmission of specific services is optimized from the network equipment aspect. Fig. 9 The process of packet control in the P4 switch is shown, where four queues correspond to three different priorities, as described below.

[0122] Queue A: provides low-latency transmission service forwarding.

[0123] Queue B: Serves real-time high-bandwidth transmission business forwarding.

[0124] C queue: used to provide management information transmission service forwarding.

[0125] D queue: The speed is limited and it is used to forward unclassified services.

[0126] The specific algorithm for data packet scheduling is as follows. In this algorithm, different priority queues are independent of each other. The high priority queue can receive service first. Only when there is no data in the high priority queue will the low priority queue be processed, otherwise it will wait. The overall processing flow is as follows: Fig.10 .

[0127] In an exemplary embodiment, step S3 specifically includes:

[0128] S31: judging whether the executor pool is empty according to the service type of the new data packet and the corresponding service demand, and obtaining a fourth judgment result.

[0129] S32: If the fourth judgment result is yes, determine whether the service queue is full to obtain a fifth judgment result.

[0130] S33: If the fifth judgment result is yes, a rejection strategy is adopted; the rejection strategy includes: directly abandoning or discarding the business at the front of the queue, and then repeatedly submitting the rejected business and having the caller execute the business.

[0131] S34: If the fifth judgment result is no, the new data packet is queued.

[0132] S35: If the fourth judgment result is no, then classify the business types.

[0133] S36: After the service type classification is completed, the services whose types have been classified are distributed to corresponding sub-executors, and it is determined whether the corresponding sub-executors are full, thereby obtaining a sixth determination result.

[0134] S37: If the sixth judgment result is yes, then wait, and if the waiting time exceeds the preset time, then throw out processing.

[0135] S38: If the sixth judgment result is no, the type-classified business is assigned to the corresponding sub-executor, and the resources are released after the completion of the waiting, so as to realize dynamic resource matching.

[0136] 4. State-based optimized routing mechanism.

[0137] In terms of heterogeneous network routing, it mainly uses network resource perception data and business perception data, and performs network-wide routing planning through algorithms running in the controller to ensure end-to-end data transmission path optimization. In the application service module, it mainly applies the functions provided in the core business module to realize the perception of business needs in the industrial Internet, accurate matching of network needs, and high-reliability and low-latency transmission tasks of remote control instructions.

[0138] First, the controller builds a link state database based on the network topology and network state perception information, including the two endpoints of each link, the link delay, maximum capacity, used bandwidth, and remaining bandwidth. Specifically, for each link , record the following information ,in and Represent the starting point and end point of the link respectively. is the link delay, is the maximum capacity of the link, is the used bandwidth of the link, is the remaining bandwidth of the link.

[0139] Assume business flow The delay requirement for the link is The bandwidth requirement is The controller needs to calculate the set of links that meet its latency requirements and define The link set is ,Right now Then, determine whether there is a link set that meets the delay requirement. is empty, a no path error is returned. If it is not empty, continue to determine whether the service has bandwidth requirements. If there is no bandwidth requirement, select the shortest path, otherwise calculate the link set that meets the bandwidth requirements. Define the link set that meets the bandwidth requirements as ,Right now . Further, it is determined whether there is a link set that meets the bandwidth requirement. If is empty, a no path error is returned. If it is not empty, the link that satisfies the minimum delay is selected as the optimal link to complete the routing task.

[0140] In the optimization model, define the variables If you select Link ,but ,otherwise The optimization goal is to minimize the total delay of the path, that is, The constraints include that each selected link must meet the bandwidth requirements, namely: .

[0141] Each selected link must meet the latency requirement, namely: .

[0142] The above content ensures the continuity of the path, that is, from the source node To the target node All nodes on the path are connected. The overall processing flow is as follows Fig.11 shown.

[0143] Step 1: The controller establishes a link state database based on the network topology and global state information, including the two endpoints of each link, the link delay, maximum capacity, used bandwidth, and remaining bandwidth.

[0144] Step 2: Intercept the service flow and obtain its port information to clarify the service's latency requirements for specific network indicators.

[0145] Step 3: The controller calculates the link set that meets its latency requirements.

[0146] Step 4: Determine whether the link set of the delay requirement is empty.

[0147] Step 5: If it is empty, return No Path Error, otherwise continue to determine whether the service requires bandwidth.

[0148] Step 6: If there is no demand for bandwidth, arrange the shortest path with the shortest number of hops; otherwise, calculate the set of links that meet the bandwidth.

[0149] Step 7: If the set is empty, a no-path Error is returned, and the link with the minimum delay is selected as the optimal link.

[0150] Step 8: Complete the route selection task.

[0151] As an optional implementation, in step S4, it specifically includes.

[0152] S41: Establish a link state database according to the network topology and the global state information of the network; the link state database includes: two end points of each link, link delay, maximum capacity, used bandwidth and remaining bandwidth.

[0153] S42: Senses the service demand of new data packets through ports.

[0154] S43: Calculate a link set that meets the delay requirement according to the link state database and the service requirement of the new data packet.

[0155] S44: Determine whether the link set that meets the delay requirement is empty, and obtain a seventh determination result.

[0156] S45: If the seventh judgment result is yes, return No Path Error.

[0157] S46: If the seventh judgment result is no, determine whether there is a demand for bandwidth to obtain an eighth judgment result.

[0158] S47: If the eighth judgment result is no, determining the link with the minimum number of hops as the optimal transmission path.

[0159] S48: If the eighth judgment result is yes, determine whether there is a path that meets the bandwidth requirement to obtain a ninth judgment result.

[0160] S49: If the ninth judgment result is no, then return No Path Error.

[0161] S410: If the ninth judgment result is yes, the link with the minimum delay is determined as the optimal transmission path.

[0162] This application proposes a deterministic network service method based on business characteristic perception, which has the following advantages compared with existing network perception and service provision methods.

[0163] 1. Actively send detection packets to the network to ensure that the detection packets are in the same path as the business flow, and flexibly analyze network behavior based on the characteristic changes of the detection packets affected by the integrated network. Indicators include network status information such as packet loss rate, delay, jitter, and available bandwidth.

[0164] 2. The task demand-driven scheduling mechanism identifies task importance indicators and combines node resources and business requests for scheduling, thereby meeting the industry's network requirements for specific tasks and improving the level of differentiated service.

[0165] 3. Based on the perception of industry network demand and signal measurement information such as network latency, through mechanisms such as deterministic routing, the end-to-end latency and other performance indicators of each service are ensured, thereby meeting the deterministic transmission quality requirements of specific network transmission services such as industrial control.

[0166] Based on the same inventive concept, the embodiment of the present application also provides a deterministic network service device based on business characteristic perception for implementing the above-mentioned. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme recorded in the above-mentioned method, so the specific limitations in the one or more embodiments of the deterministic network service device based on business characteristic perception provided below can refer to the limitations of the deterministic network service method based on business characteristic perception in the above text, and will not be repeated here.

[0167] In an exemplary embodiment, Fig.12 As shown, a deterministic network service device based on business characteristic perception is provided, including:

[0168] The information acquisition module M1 is used to acquire the global status information of the network; the global status information includes: packet loss rate, delay, jitter and available bandwidth.

[0169] The service type and service requirement determination module M2 is used to determine the service type and corresponding service requirement of a new data packet when a new data packet needs to be transmitted in the network.

[0170] The resource dynamic matching module M3 is used to dynamically match resources according to the service type of the new data packet and the corresponding service demand.

[0171] The optimal transmission path determination module M4 is used to calculate the optimal transmission path according to the global state information of the network when the current network has sufficient capacity to support the service demand of the new data packet.

[0172] The allocation module M5 is used to allocate the new data packet to a corresponding forwarding queue according to the service type of the new data packet and the optimal transmission path.

[0173] This application can enable the network status to be transmitted along the same path as the service flow data packets in the programmable switching device area, ensuring the real-time and accuracy of network status acquisition. In addition, a real-time network perception solution compatible with traditional network equipment is designed to protect existing investments and support network perception. At the same time, it can match the queue scheduling measurement inside the network node with the data packet transmission requirements, and optimize the transmission of characteristic services from the network equipment aspect. In addition, multi-dimensional heterogeneous network routing optimization ensures the best transmission path and resource allocation, meeting the requirements of latency and bandwidth for specific network usage scenarios.

[0174] In an exemplary embodiment, a computer device is provided. The computer device may be a server or a terminal. The internal structure diagram thereof may be as follows: Fig.13 As shown. The computer device includes a processor, a memory, an input / output interface (Input / Output, referred to as I / O) and a communication interface. The processor, the memory and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store global state information of the network. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a deterministic network service method based on business feature perception is implemented.

[0175] Those skilled in the art will understand that Fig.13 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0176] In an exemplary embodiment, a computer device is further provided, including a memory and a processor, wherein a computer program is stored in the memory, and the above-mentioned method embodiments are implemented when the processor executes the computer program.

[0177] In an exemplary embodiment, a computer-readable storage medium is provided, storing a computer program, and when the computer program is executed by a processor, the above-mentioned method embodiments are implemented.

[0178] In an exemplary embodiment, a computer program product is provided, including a computer program, and when the computer program is executed by a processor, the above-mentioned method embodiments are implemented.

[0179] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0180] This article uses specific examples to illustrate the principles and implementation methods of this application. The description of the above embodiments is only used to help understand the method and core ideas of this application. At the same time, for those skilled in the art, according to the ideas of this application, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting this application.

Claims

1. A deterministic network service method based on business characteristic perception, characterized in that: The deterministic network service method based on business characteristic perception includes: Obtaining global status information of the network; the global status information includes: packet loss rate, delay, jitter and available bandwidth; When a new data packet needs to be transmitted in the network, determine the service type and corresponding service requirements of the new data packet; Dynamically matching resources according to the service type and corresponding service requirements of the new data packet; When the current network has sufficient capacity to support the service demand of the new data packet, the optimal transmission path is calculated according to the global state information of the network; Allocating the new data packet to a corresponding forwarding queue according to the service type of the new data packet and the optimal transmission path; Get the global status information of the network, including: Construct a probe packet; Forwarding the detection data packet across the entire network to obtain global status information of the network; The detection data packet is forwarded throughout the network to obtain the global status information of the network, including: When the detection data packet reaches the first switch, determining whether the first switch supports programmability, and obtaining a first determination result; If the first judgment result is yes, mirroring the detection data packet to obtain a mirrored original message; Inserting the required network information into the mirrored original message and forwarding it to the next-hop switch; the required network information is metadata encapsulated from the telemetry information specified by the detection header; When the next-hop switch matches the detection header, the required network information is inserted, and it is determined whether the last-hop switch is reached to obtain a second determination result; If the second judgment result is no, return to the step of "forwarding to the next hop switch"; If the second judgment result is yes, then judging whether the last hop switch supports programmability, and obtaining a third judgment result; If the third judgment result is yes, the last hop switch inserts the required network information after matching the detection header, and encapsulates an IP header; the IP is the telemetry server address; If the third judgment result is no, directly encapsulate the IP header; If the first judgment result is no, traditional IP forwarding is adopted, and the process returns to the step of "forwarding to the next hop switch"; Parsing telemetry information according to the IP header to obtain global status information of the network and reporting it to an upper-layer telemetry application; Dynamically matching resources according to the service type of the new data packet and the corresponding service requirements, specifically including: According to the service type and corresponding service requirements of the new data packet, determining whether the executor pool is empty, and obtaining a fourth determination result; If the fourth judgment result is yes, then judging whether the service queue is full, and obtaining a fifth judgment result; If the fifth judgment result is yes, a rejection strategy is adopted; the rejection strategy includes: directly abandoning or discarding the business at the front of the queue, and then repeatedly submitting the rejected business and having the caller execute the business; If the fifth judgment result is no, the new data packet is queued; If the fourth judgment result is no, then classifying the business types; When the business type classification is completed, the business with the type classification completed is distributed to the corresponding sub-executor, and it is determined whether the corresponding sub-executor is full, to obtain a sixth determination result; If the sixth judgment result is yes, then wait, if the waiting time exceeds the predetermined time, then throw out processing; If the sixth judgment result is no, the type-classified business is assigned to the corresponding sub-executor, and the resources are released after waiting for completion, so as to realize dynamic resource matching.

2. The deterministic network service method based on business characteristic perception according to claim 1 is characterized in that: When a new data packet needs to be transmitted in the network, the service type and corresponding service requirements of the new data packet are determined, including: Parse the new data packet according to the port to obtain the service type of the new data packet; According to the service type of the new data packet, the corresponding specific requirement is determined.

3. The deterministic network service method based on business characteristic perception according to claim 1, characterized in that: When the current network has sufficient capacity to support the service requirements of the new data packet, the optimal transmission path is calculated based on the global state information of the network, specifically including: Establish a link state database based on the network topology and the global state information of the network; the link state database includes: the two end points of each link, the link delay, the maximum capacity, the used bandwidth and the remaining bandwidth; Sense the business needs of new data packets through ports; Calculating a link set that meets the delay requirement according to the link state database and the service requirement of the new data packet; Determine whether the link set that meets the delay requirement is empty, and obtain a seventh determination result; If the seventh judgment result is yes, then return No Path Error; If the seventh judgment result is no, then judging whether there is a demand for bandwidth, and obtaining an eighth judgment result; If the eighth judgment result is no, determining the link with the minimum number of hops as the optimal transmission path; If the eighth judgment result is yes, then judging whether there is a path that meets the bandwidth requirement, and obtaining a ninth judgment result; If the ninth judgment result is no, then return No Path Error; If the ninth judgment result is yes, the link with the minimum delay is determined as the optimal transmission path.

4. A deterministic network service device based on business characteristic perception, the device implements the deterministic network service method based on business characteristic perception according to any one of claims 1 to 3, characterized in that: The deterministic network service device based on business characteristic perception includes: An information acquisition module, used to acquire global status information of the network; the global status information includes: packet loss rate, delay, jitter and available bandwidth; A service type and service requirement determination module, used to determine the service type and corresponding service requirement of a new data packet when a new data packet needs to be transmitted in the network; A resource dynamic matching module, used for dynamically matching resources according to the service type and corresponding service requirements of the new data packet; An optimal transmission path determination module, used to calculate the optimal transmission path according to the global state information of the network when the current network has sufficient capacity to support the business needs of the new data packet; The allocation module is used to allocate the new data packet to a corresponding forwarding queue according to the service type of the new data packet and the optimal transmission path.

5. A computer device comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the deterministic network service method based on business characteristic perception as described in any one of claims 1 to 3.

6. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the deterministic network service method based on business characteristic perception described in any one of claims 1 to 3 is implemented.

7. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the deterministic network service method based on business characteristic perception described in any one of claims 1 to 3 is implemented.

Images