Parameter configuration method for burst degree control and time delay guarantee of dynamic load network

Abstract

The invention discloses a burst degree control and delay guarantee parameter configuration method for a time-sensitive network, which comprises the following steps of: directly constraining an accumulated upper limit of a credit value in a CBS mechanism by introducing a credit upper limit parameter CLP which is independent of a credit growth rate and can be dynamically adjusted, and further limiting excessive occupation of burst flow on channel resources from the source; meanwhile, a worst-case delay analysis model of an integrated credit upper bound CLP is constructed based on a network calculation theory, end-to-end delay constraints of traffic can still be strictly met while traffic burst control is implemented, and burst control and delay guarantee under a dynamic load network are realized. The method has the advantages that through cooperative regulation and control of a single additional parameter CLP, unification of burst control and deterministic time delay guarantee is achieved on the premise that existing CBS standardized parameter configuration is not changed, and the performance and deployment flexibility of the TSN in a dynamic load environment are remarkably improved.

Description

Technical Field

[0001] This invention relates to the technical field of CBS parameter configuration in Time-Sensitive Networks (TSNs), and more particularly, to a method for configuring burst control (BC) and delay guarantee (DG) parameters based on a credit limit parameter (CLP) under dynamic load conditions. Background Technology

[0002] Time-Sensitive Networking (TSN), with its core characteristics of deterministic low latency, microsecond-level clock synchronization, and highly reliable data transmission, has become a core communication technology in safety-critical fields such as the next-generation Industrial Internet, Vehicular Networks (IVN), and avionics systems. The TSN standard system, through a series of traffic scheduling, clock synchronization, and reliability enhancement mechanisms, aims to achieve the converged transmission of heterogeneous traffic based on the general-purpose Ethernet technology platform, while providing bounded and predictable end-to-end performance guarantees for critical business traffic. In this context, ensuring that high-priority traffic remains unaffected by background traffic interference in complex dynamic load environments, and consistently meets its stringent latency and jitter limits, has become a core challenge for the implementation of TSN technology.

[0003] In the traffic shaping mechanism of TSN, the Credit-Based Shaper (CBS), as the core shaping algorithm defined by the IEEE 802.1Qav standard, is widely used to ensure the determinism of traffic transmission behavior and plays a crucial role in ensuring real-time traffic performance. The core idea of ​​the CBS mechanism is to dynamically manage the independent virtual credit account of each queue, thereby achieving smooth traffic shaping. Its basic operating rules can be summarized as follows: when there are data frames waiting to be sent in the queue, if the credit value is non-negative and the channel is idle, transmission is allowed, and the credit amount decreases at the rate of sendSlope during transmission; when the queue is empty and the credit value is negative, the credit amount recovers to 0 at the rate of idleSlope; when the queue cannot send due to waiting for the channel to become idle, the credit amount accumulates at the rate of idleSlope. The CBS shaping mechanism only allows the transmission of data packets for the corresponding queue if and only if the credit value of the traffic queue is non-negative. This transmission method, which dynamically increases or decreases the credit value, effectively smooths the traffic transmitted in the network into an approximately constant data stream, and achieves bandwidth isolation between multiple queues. This prevents high-priority traffic from excessively occupying the channel and starving low-priority traffic, providing a solid deterministic latency foundation for services such as Audio / Video Bridging (AVB).

[0004] The parameter configuration mechanism of the CBS shaper reveals its insufficient adaptability when facing unpredictable burst traffic in modern networks, potentially leading to issues such as timeouts, packet loss, and buffer overflows. In dynamic load environments, configuring only the sendSlope and idleSlope of credits without constraining the upper limit of credit growth means that the idleSlope parameter cannot directly and temporarily control the burst traffic behavior. This allows data frames in high-burst queues to continuously occupy the channel and be transmitted for a considerable period, creating intense channel resource competition with other normally scheduled traffic in the network. This significantly increases the transmission latency and jitter of other traffic queues, undermining the deterministic guarantee of TSN network transmission. Furthermore, burst traffic can also cause momentary buffer overflows at network nodes, leading to packet loss. This not only threatens transmission reliability but also further exacerbates network load through retransmission mechanisms, creating a vicious cycle. The root cause is the lack of effective control and constraints on the upper bound of credit, which prevents the CBS mechanism from restraining its continuous channel occupation behavior in the event of sudden traffic, thus leading to the destruction of the overall deterministic latency guarantee of the network.

[0005] Current optimization research on CBS mechanisms mainly focuses on the configuration optimization of static parameters, such as determining the idleSlope parameter of each queue through offline calculation. While these methods can provide certain performance guarantees under fixed business modes, they lack adaptability to dynamic business scenarios. When unexpected bursts of traffic occur in the network, existing solutions often require reconfiguring the CBS parameters of all nodes in the network. This process not only introduces significant management overhead but may also cause interruptions in service transmission, making it difficult to meet the dual requirements of network systems for flexibility and real-time performance. Summary of the Invention

[0006] To address the limitations of the Credit Growth Rate (CBS) mechanism in handling dynamic load network environments, this invention proposes a burst control and delay guarantee parameter configuration method based on Credit Limit (CLP) in dynamic load networks. This method primarily addresses the technical challenge of existing CBS mechanisms, where standardized credit growth rate parameters cannot directly control the transmission behavior of burst traffic in scenarios with short-term concentrated bursts of traffic. This leads to burst traffic continuously occupying channel resources, thereby compromising the deterministic delay guarantee of other traffic. By introducing a credit limit parameter and establishing a corresponding end-to-end worst-case delay assessment model, this invention effectively controls traffic burstiness while satisfying end-to-end delay constraints, achieving an optimal balance between burst control and delay guarantees. This invention constructs a worst-case delay assessment model containing a credit limit based on network calculus theory. Compared to existing delay assessment models, this model introduces the credit growth limit parameter CLP. Through dynamic adjustment of this parameter, it effectively constrains the burstiness of traffic while strictly ensuring its real-time performance, thus achieving an optimal balance between burst traffic control and deterministic delay guarantees. The core idea of ​​this method is to constrain the upper limit of credit accumulation through CLP parameters, thereby limiting the bandwidth available to traffic and preventing burst traffic from excessively occupying channel resources.

[0007] Credit-based shaper (CBS) has become one of the core mechanisms of time-sensitive networks, achieving low-jitter transmission of service traffic through bandwidth isolation and deterministic delay guarantees. However, in dynamic load environments, the standardized idleSlope parameter configuration of the CBS mechanism is difficult to adaptively adjust when faced with short-term concentrated injections of burst traffic, failing to directly and quickly control the burst behavior of traffic, leading to queue congestion and affecting the deterministic delay guarantees of other service traffic. To address the problems of the CBS mechanism in handling dynamic load network environments, this invention proposes a burst control and delay guarantee parameter configuration method based on credit upper bound (CLP), and constructs a worst-case delay evaluation model containing the credit upper bound based on network calculus theory, providing a direct and efficient method for CBS to cope with burst traffic challenges. This method constrains credit growth through the credit upper bound (CLP) parameter, effectively controlling traffic burstiness while strictly satisfying end-to-end delay constraints. Specifically, this invention first analyzes the target network topology, traffic characteristics, and CBS credit parameter configuration information. Building upon this foundation, a joint latency-burst assessment model based on the upper bound of Credit Limit (CLP) is established. For each type of traffic, the CLP is configured as a specific percentage of the theoretical maximum value of the credit limit, according to its end-to-end latency constraints and burst requirements. This constructs the traffic arrival and service curves at network nodes constrained by the CLP. Subsequently, the maximum burst and maximum latency of traffic at each node are calculated based on this model, and their compliance with the deterministic latency guarantee required by the business is verified. By iteratively adjusting the CLP parameters and repeating the evaluation process, the optimal CLP configuration that minimizes traffic burst is finally obtained while satisfying the latency constraints.

[0008] The method proposed in this invention can overcome the limitations of the CBS mechanism in dynamic load environments, reduce the interference of burst traffic on coexisting traffic, and achieve dynamic balance between burstability and latency guarantee without reconfiguring the idleSlope parameter. Furthermore, by continuously monitoring the network operating status and re-triggering the CLP optimization process when topology or traffic characteristics change, the network's adaptability and deployment flexibility in dynamic load environments are further enhanced. The burstability control and latency guarantee parameter configuration method based on Credit Upper Bound (CLP) proposed in this invention includes the following steps:

[0009] Step 1: Analyze the information of the target network;

[0010] Step 101: Obtain the topology of the target network and parse the connection relationships between nodes;

[0011] Step 102: Obtain all traffic information transmitted in the network and analyze the traffic characteristics;

[0012] Step 103: Obtain the parameter configuration information of the credit-based shaper in the network;

[0013] Step 2: Establish a latency-burst assessment model based on the upper bound of credit volume;

[0014] Step 201: For each type of traffic, configure the credit growth cap (CLP) to a specific percentage of the theoretical maximum value based on its end-to-end latency constraints and burst requirements.

[0015] Step 202: Calculate the traffic arrival curve at the network node;

[0016] Step 203: Calculate the service curves at network nodes after CLP constraints;

[0017] Step 3: Calculate and verify the burstiness and latency guarantee of the traffic;

[0018] Step 301: Based on the constrained arrival and service curves, calculate the maximum burstiness and maximum latency of traffic at a single node;

[0019] Step 302: Compare the calculated end-to-end latency with the requirements of the service flow to verify whether it meets the deterministic latency guarantee.

[0020] Step 303: Iteratively adjust the CLP parameters and jump to step 201 until the optimal CLP value that minimizes traffic burstiness is found while satisfying the end-to-end delay constraint.

[0021] Step 304: Monitor the network operation status. If the traffic characteristics or network topology change, jump to step one and set the CLP parameters for the next round; otherwise, output the optimal CLP configuration value.

[0022] The technical effects achieved by the method of this invention are as follows:

[0023] (1) By introducing a credit limit parameter CLP that is independent of the credit growth rate and can be dynamically adjusted, this invention effectively solves the core technical problem that existing credit-based shapers may cause excessive occupation of channel resources by sudden traffic in dynamic load environments, which may undermine the deterministic delay guarantee of other service flows in the network. It also solves the problem that the CBS mechanism is not adaptable to sudden traffic in dynamic load network environments.

[0024] (2) Based on network calculus theory, this invention constructs a worst-case delay assessment model with a credit limit CLP. By obtaining the delay boundary of traffic transmission, the model can provide a quantifiable and accurate performance boundary for traffic, thus providing a solid theoretical basis for the optimal configuration of CLP parameters and achieving the best balance between traffic burst control and delay guarantee.

[0025] (3) This invention achieves the unity of burst control and deterministic delay guarantee without changing the existing CBS standardized parameter configuration through the coordinated regulation of a single additional parameter CLP, which significantly improves the performance and deployment flexibility of TSN in dynamic load environments.

[0026] (4) By continuously monitoring changes in network topology and traffic characteristics, the present invention can recalculate and optimize CLP parameters, enabling the network to adapt to environmental changes and maintain optimal performance. Attached Figure Description

[0027] Figure 1 This is a flowchart of the network traffic burst control and delay guarantee method based on the upper bound of credit volume according to the present invention.

[0028] Figure 2 This is a worst-case latency derivation graph for traffic under the CBS shaping mechanism.

[0029] Figure 3 This is a schematic diagram of the topology of an example TSN network.

[0030] Figure 4 (A) is a credit change curve without the introduction of the credit growth cap parameter CLP; (B) is a credit change curve with the introduction of the credit growth cap parameter CLP.

[0031] Figure 5 This invention introduces a credit growth cap parameter CLP backend-to-end latency variation curve.

[0032] Figure 6 This is a curve showing the change in suddenness after the introduction of the credit growth cap parameter CLP in this invention. Detailed Implementation

[0033] The present invention will now be described in further detail with reference to the accompanying drawings and embodiments. The examples of the parameters listed are merely preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

[0034] In this invention, the network topology used by the TSN network is denoted as... And network topology It contains a set of nodes and Connecting Edge Set ,Right now .

[0035] subscript The subscript indicates the identifier of a node in the network topology. This represents the total number of nodes in the network topology, where: This represents the first node. This indicates the second node. Indicates the first Each node. This represents the last node. For ease of explanation, Also known as any node.

[0036] In this invention, devices such as end systems (ES) or switches (SW) have multiple interfaces (or ports, which can be numbered and labeled). For example, an interface used for signal input is called an input port, and another interface used for signal output is called an output port. Alternatively, any node can be... The input port is denoted as any node The output port is denoted as .

[0037] In this invention, any node in the network topology is used. Abstracted as an end system or switch Nodes can be distinguished using consecutive numerical codes, such as Figure 3 As shown. In the network topology, directed connections are represented using a directed connection matrix, denoted as . The A value of 1 indicates the existence of a directed edge, while a value of 0 indicates no directed edge. By... The value of is used to indicate whether a directed edge exists between two nodes. The set of links connected to an edge node is . .

[0038] The link bandwidth of a TSN network is denoted as C.

[0039] In network topology The set of network traffic in the data is denoted as ,and subscript The identifier representing traffic flow, subscript This represents the total traffic volume. Among them: This indicates the first network traffic entry. This indicates the second network traffic. Indicates the first A number of network traffic. This indicates the last network traffic entry. For ease of explanation, Also known as any network traffic.

[0040] In this invention, any network traffic The traffic characteristic information is denoted as ,and .

[0041] For traffic The length is expressed in bits.

[0042] For traffic The period for generating frames, in μs.

[0043] For traffic The type.

[0044] This invention classifies traffic in a TSN network into three types based on the varying degrees of network traffic demand for network service quality. The order of traffic importance from highest to lowest is as follows: , and ,in , Type traffic passes through the CBS shaper, Type of traffic is background interference traffic and does not pass through the CBS shaper.

[0045] For traffic The transmission path can be represented by a set of nodes along the way. transmission path , recorded as , where nodes Marked as the current node, node It is a node The previous node, node It is a node The next node, and the transmission path There are N nodes in total.

[0046] Traffic rate, which is the bandwidth used for traffic transmission, is measured in units of bandwidth. ,and .

[0047] For flow along Transmission, the maximum burst level at each node, measured in bits. The burst level at each node along the path can be represented using numbered information including the node numbers (excluding the destination node). For example, traffic. The nodes traversed from the start node to the end node are: , , , , , Then the flow The suddenness statistics on the path are as follows .

[0048] This refers to the delay at each node along the path of traffic as it travels from the first node to the last, often simply called the transmission delay along a similar traffic path, measured in μs. The delay value at each node along the path can be represented using numbered information including the node numbers (excluding the destination node). For example, traffic... The nodes traversed from the start node to the end node are: , , , , , Then the flow The transmission delay statistics on the path are as follows .

[0049] For traffic The end-to-end transmission deadline, in units of , represents the sum of the delay values ​​at each node along the route, and .

[0050] Worst latency of traffic

[0051] like Figure 2 As shown, in this invention, the arrival curve of the traffic is a (rate, burst) sloping line model, where rate refers to the arrival speed of the curve, represented by the slope of the straight line, and burst refers to the burstiness, represented by the vertical intercept of the straight line. The service curve of the output port is a (rate, time) sloping line model, where rate refers to the service speed of the curve, represented by the slope of the straight line, and time refers to the initial delay, represented by the horizontal intercept of the straight line. When the arrival curve of the traffic and the service curve of the output port are plotted on the same graph, the maximum horizontal intercept of the two curves represents the traffic volume. The worst-case latency at that port.

[0052] The process of calculating the end-to-end latency of traffic is progressive, calculating the burstiness of any traffic flow based on the current output port's traffic. The maximum latency value, and then the traffic is calculated based on the latency value. The burstiness of the output to the next node is used as the basis for calculating the latency of the traffic at the next output port. After calculating the worst-case maximum latency for each type of traffic at all output ports, the latency of any traffic will be obtained. Along its transmission path The sum of the delay values ​​corresponding to the output ports is the flow rate. End-to-end maximum delay The calculation formula is: .

[0053] See Figure 1 As shown, a parameter configuration method for burst degree control and delay guarantee in dynamic load networks according to the present invention includes the following steps:

[0054] Step 1: Analyze the information of the target network;

[0055] This invention achieves comprehensive information analysis of the target TSN network through step one. Specifically, firstly, the topology of the target network is acquired and analyzed to clarify the connection relationships between network nodes; secondly, all traffic information transmitted in the network is acquired, and traffic characteristics such as period, frame length, and type are analyzed; finally, the parameter configurations of all credit-based shapers in the network are acquired. The purpose of this step is to lay the foundation for establishing an accurate evaluation model, ensuring that the model input matches the actual network structure.

[0056] Step 101: Obtain the target network topology and parse the connection relationships between nodes;

[0057] In this invention, the TSN network topology is denoted as... The set of network traffic is denoted as Any traffic The traffic characteristic information is denoted as .

[0058] Network topology It contains node set information and edge set information Abstracting nodes in the TSN network into end systems. and switches Furthermore, nodes are numbered consecutively. The network topology uses a directed connection matrix to represent directed edges, denoted as . , , The , , A value of 1 indicates the existence of a directed edge, while a value of 0 indicates no directed edge. By... , , The assignment is used to represent the relationship between two nodes (end system). to switch ,switch to switch ,switch End-to-end system Does a directed edge exist?

[0059] Step 102: Obtain all traffic information transmitted in the network and analyze the traffic characteristics;

[0060] Any traffic Since it contains various feature information, this invention uses a structure to construct traffic feature information. ,and .

[0061] Similarly, we can deduce that traffic Traffic characteristic information ,and .

[0062] Similarly, we can deduce that traffic Traffic characteristic information ,and .

[0063] Similarly, we can deduce that traffic Traffic characteristic information ,and .

[0064] Similarly, we can deduce that traffic Traffic characteristic information ,and .

[0065] Then we have: Flow rate – Flow rate feature set .

[0066] Step 103: Obtain the parameter configuration information of the credit-based shaper in the network;

[0067] Obtain the integer parameters set for the CBS queue in the parameter configuration of each node device (such as the end system ES or switch SW in Figure 3), including the credit growth rate. With the rate of credit reduction .in, Indicates flow rate of Attributes are The rate of credit growth Indicates flow rate of Attributes are The rate of credit reduction. In this invention, , Where C represents the link bandwidth of the TSN network.

[0068] Step 2: Establish a latency-burst assessment model based on the upper bound of credit volume;

[0069] For ease of explanation, when performing parameterization based on the upper bound of the credit limit parameter (CLP), we will denote it as... .

[0070] This invention establishes a latency-burst assessment model based on an upper bound of credit value in step two. First, for each type of traffic, the credit growth cap of CBS is configured as a specific percentage of the theoretical maximum value. Second, based on this configuration, the arrival curve of the traffic at each network node is calculated. Simultaneously, the service curve that the network node can provide for that type of traffic, constrained by the credit growth cap parameter CLP, is calculated. The purpose of this step is to calculate the arrival and service curves required for network computation, providing a foundation for calculating the end-to-end latency of the traffic in step three.

[0071] Step 201: For each type of traffic, configure the credit growth cap (CLP) to a specific percentage of the theoretical maximum value based on its end-to-end latency constraints and burst requirements.

[0072] In this invention, the definition is... This is the lower bound of the credit volume. This represents the upper limit of credit volume. This is the upper limit for credit growth. express Attributes are Type of traffic on port The lower bound of credit volume. express Attributes are Type of traffic on port The upper limit of credit volume, express Attributes are Type of traffic on port The upper limit of credit growth, and .set up for The percentage, that is .

[0073] Step 202: Calculate the traffic arrival curve at the network node;

[0074] In this invention, the definition is... For the total flow rate, For sudden traffic bursts, For flow through the port Maximum frame length, The calculation formula is ,in For flow through the port The collection of all traffic. For all ports ,and Attributes are A collection of traffic types. express Attributes are Type of traffic on port Total flow rate The calculation formula is: . express Attributes are Type of traffic on port Traffic bursts The calculation formula is: .

[0075] according to and It can be obtained Attributes are Type of traffic on port arrival curve , The calculation formula is Where C represents the link bandwidth of the TSN network, and t represents the traffic transmission time.

[0076] Attributes are Type of traffic on port arrival curve The end-to-end delay of the flow is calculated in step 301.

[0077] Step 203: Calculate the service curves at network nodes after CLP constraints;

[0078] definition For traffic service rate, The initial service delay for the port. express Attributes are Type of traffic on port The traffic service rate is equal to the rate provided by the CBS mechanism. Credit growth rate of traffic type configuration . express In the flow Attributes are Type of traffic on port Initial service delay.

[0079] In this invention, the recipient Traffic service rate after parameter constraints The calculation formula is: Where C represents the link bandwidth of the TSN network. Initial service delay of the port after parameter constraints The calculation formula is: .

[0080] according to and It can be obtained Attributes are Type of traffic on port service curve , The calculation formula is , where t represents the data transmission time.

[0081] Attributes are Type of traffic on port service curve The end-to-end delay of the flow is calculated in step 301.

[0082] Step 3: Calculate and verify the burstiness and latency guarantee of the traffic;

[0083] This invention optimizes the calculation and verification of traffic burstability and latency guarantees through step three. First, based on the arrival and service curves obtained in step two, the maximum burstability and maximum latency of traffic at each single node are calculated. Then, the latency of each node is accumulated to obtain the end-to-end latency, which is compared and verified against the deterministic latency requirements of the service flow. By iteratively adjusting the credit growth upper limit parameter and repeating the above calculation process, the optimal CLP parameter value is finally found that minimizes traffic burstability and meets the deadline while satisfying the end-to-end latency constraint. Furthermore, this invention continuously monitors the network operating status, and further configures the CLP parameters once changes in traffic characteristics and / or network topology are detected. The purpose of this step is to find the optimal solution for network configuration, effectively suppressing traffic burstability while strictly guaranteeing deterministic latency, thereby improving the flexibility and stability of the network.

[0084] Step 301: Based on the constrained arrival and service curves, calculate the maximum burstiness and maximum latency of traffic at a single node;

[0085] (a) Maximum burst

[0086] Burst rate refers to the maximum amount of data that a queue can continuously send, defined as follows: In this invention, the recipient Port after parameter constraints middle Attributes are Burst of traffic type The calculation formula is ,in For flow through the port Maximum frame length.

[0087] In this invention, if any flow of Attributes are When the type is used, the flow rate is high. At the port suddenness equal The calculation formula is: ,in For flow through the port Maximum frame length.

[0088] (b) Maximum delay

[0089] In this invention, flow rate any transmission path Recorded as , where nodes Marked as the current node, node It is a node The previous node, node It is a node The next node.

[0090] Based on the flow arrival curves of the node output ports obtained in steps 202 and 203 and node service curve The port can be calculated. middle Attributes are Maximum evaluation latency for this type of traffic The calculation formula is: .

[0091] In network calculus, the maximum horizontal intercept used to represent the arrival curve and the service curve is considered to be the maximum latency value.

[0092] In this invention, if any flow of Attributes are When the type is used, the flow rate is high. At the port Maximum evaluation delay equal The calculation formula is: .

[0093] Step 302: Compare the calculated end-to-end latency with the requirements of the service flow to verify whether it meets the deterministic latency guarantee.

[0094] In this invention, for each traffic... According to traffic transmission path All nodes that the transmission passes through suddenness By comparing and taking the larger value, we can obtain the traffic. In the transmission path Maximum burst The calculation formula is: .

[0095] In this invention, for each traffic... According to traffic transmission path All nodes that the transmission passes through Maximum delay By summing, we can obtain the flow rate. In the transmission path End-to-end maximum delay The calculation formula is: .

[0096] Based on the calculated end-to-end delay of the traffic, determine any traffic... Does it meet the transmission objective of deterministic delay guarantee, that is, in terms of determinism, for any current traffic... The transmission time is less than the end-to-end transmission deadline of the traffic.

[0097] Step 303: Iteratively adjust the CLP parameters and jump to step 201 until the optimal CLP value that minimizes traffic burstiness is found while satisfying the end-to-end delay constraint.

[0098] Based on the calculation results of the deterministic delay guarantee in step 302, update The size. If the calculation result of step 302 meets the end-to-end delay requirement, then appropriately reduce it in the next iteration. The parameter value is adjusted to achieve a smaller burstiness; if the calculation result of step 302 does not meet the end-to-end delay requirement, the value is appropriately increased in the next iteration. The parameter value is used to satisfy any traffic. The latency requirement. If any traffic is satisfied... Under the premise of end-to-end latency requirements, The minimum value has been obtained, that is, the minimum value has been found. If the optimal value is found, then proceed to step 304.

[0099] In this invention, by jumping to step 201, it is ensured that, under the premise of strictly satisfying all traffic deterministic delay guarantees, the optimal method that minimizes traffic burstiness is successfully found. Configure values ​​to achieve the best balance between emergency control and delay protection.

[0100] Step 304: Monitor the network operation status. If the traffic characteristics or network topology change, jump to step one and set the CLP parameters for the next round; otherwise, output the optimal CLP configuration value.

[0101] In step 304, the present invention will continuously monitor the network's operating status. If the traffic characteristics or network topology of the TSN network change, and the burstiness requirements and deterministic latency guarantees of the traffic in the TSN network cannot be met, then the process will jump to step one to set the next round of CLP parameters, find the optimal CLP configuration value, and achieve the best balance between burstiness control and latency guarantee. Example

[0102] Figure 3 The diagram shows a topology of a TSN network consisting of 4 switches and 5 end systems.

[0103] The link bandwidth of the TSN network is set to 100. ,Right now .

[0104] The credit growth rate parameters for various types of traffic on the TSN network are set to...

[0105] in accordance with Figure 3 Transmitted in the TSN network of the scene , , The traffic attribute settings for each type of traffic are shown in Table 1 below. The transmission path for each type of traffic is set to the shortest transmission path.

[0106] Table 1 In Example 1 , , Type of traffic attribute configuration parameters

[0107]

[0108] In Example 1, the deadline for each flow is... Equal to each flow Configure according to the traffic settings. .

[0109] In Example 1, 48 traffic flows are first configured in the TSN network as a baseline load; subsequently, to simulate a dynamic load environment and test the system's resilience to bursts, an additional 48 burst traffic flows are injected as disturbances, bringing the total network traffic to 96 flows. Furthermore, in this invention, it is considered that... The traffic is background interference traffic, and there is no need to consider the end-to-end latency and burstiness of this type of traffic.

[0110] Based on the parameter configuration of the TSN network, and using a latency-burst assessment model with an upper bound on the credit limit, the maximum end-to-end latency and maximum burst rate of various traffic types before and after the presence or absence of CLP parameters are calculated.

[0111] Table 2. Calculation results of the latency-burst assessment model based on the upper bound of credit volume in Example 1.

[0112]

[0113] This example illustrates the changes in credit after introducing the Credit Growth Cap parameter (CLP). Assume that any node in the TSN network transmits various types of traffic as follows: Figure 4 As shown. Figure 4 This diagram illustrates the dynamic changes in credit for various traffic types within any node of the TSN network after introducing the Credit Growth Cap (CLP) parameter. For ease of description, A, B, and C are used to represent... , , Traffic types are represented, and their credit value changes during transmission are indicated by curves of different colors. Figure 4 (a) shows the credit change curve of traffic when CLP is not set. Figure 4 (b) is a graph showing the change in traffic credit after introducing CLP under the same transmission scenario. In this invention, it is considered that... This type of traffic is background interference traffic, and there is no need to consider the credit changes of this type of traffic.

[0114] from Figure 4 It can be clearly seen in (b) that when , When the credit value of a traffic type grows to the set CLP threshold, its credit value stops accumulating and is capped at the CLP value until the credit decreases. The credit growth cap parameter effectively constrains the upper limit of credit accumulation for the corresponding traffic category, thereby limiting the bandwidth available for traffic and preventing burst traffic from excessively consuming channel resources.

[0115] Figure 5 , Figure 6 This demonstrates the network performance changes after introducing the Credit Growth Cap parameter CLP in Example 1. Among them, Figure 5 The end-to-end delay variation curve after introducing CLP, Figure 6 The graph shows the traffic burst rate variation curves after the introduction of CLP. Solid lines represent end-to-end delay and burst rate without CLP, while dashed lines represent the corresponding performance metrics after the introduction of CLP. The markers on each curve correspond to each traffic stream transmitted in Example 1.

[0116] from Figure 5 , Figure 6 It can be seen that after introducing the Credit Growth Limit (CLP) parameter, the end-to-end latency of all traffic increases to some extent; at the same time, the burstiness of each traffic stream shows a decreasing trend. This result indicates that although the introduction of CLP increases transmission latency to some extent, it effectively suppresses the burstiness of traffic.

[0117] In summary, when setting Any traffic The maximum end-to-end delay is less than the end-to-end deadline. Therefore, in the next iteration, the size should be appropriately reduced. , The parameter value is adjusted to achieve a smaller burst.

Claims

1. A parameter configuration method for burst control and delay assurance in dynamic load networks, which is based on the parameter configuration of a credit shaper in the traffic shaping mechanism of time-sensitive networks; characterized in that... The steps are as follows: Step 1: Analyze the information of the target network; Step 101: Obtain the target network topology and parse the connection relationships between nodes; Let the TSN network topology be denoted as The It contains node set information and edge set information ; Let the set of network traffic in the TSN network topology be denoted as . subscript The identifier representing traffic flow, subscript This represents the total number of traffic entries; any single traffic entry is represented by... This indicates that the traffic volume The traffic characteristic information is denoted as ; Step 102: Obtain all traffic information transmitted in the network and analyze the traffic characteristics; Any traffic The included traffic characteristic information is ,and ; Similarly, we can deduce that traffic The traffic characteristic information is ,and ; Similarly, we can deduce that traffic The traffic characteristic information is ,and ; Similarly, we can deduce that traffic The traffic characteristic information is ,and ; Similarly, we can deduce that traffic The traffic characteristic information is ,and ; By analyzing the traffic characteristics of each traffic stream, we have: Traffic Stream – Traffic Feature Set ; Step 103: Obtain the parameter configuration information of the credit-based shaper in the network; Retrieve the integer parameters set for the CBS queue in the node's parameter configuration, including the credit growth rate. With the rate of credit reduction ;in, Indicates flow rate of Attributes are The rate of credit growth Indicates flow rate of Attributes are The rate of credit reduction; , Where C represents the link bandwidth of the TSN network; Step 2: Establish a latency-burst assessment model based on the upper bound of credit volume; Step 201: For each type of traffic, configure the credit growth cap (CLP) to a specific percentage of the theoretical maximum value based on its end-to-end latency constraints and burst requirements. definition This is the lower bound of the credit volume. This represents the upper limit of credit volume. This is the upper limit for credit growth; express Attributes are Type of traffic on port The lower bound of credit volume; express Attributes are Type of traffic on port The upper limit of credit volume; express Attributes are Type of traffic on port The upper limit of credit growth, and ; set up for The percentage, that is ; Step 202: Calculate the traffic arrival curve at the network node; definition For the total flow rate, For sudden traffic bursts, For flow through the port The maximum frame length, and The calculation formula is ; in For flow through the port The collection of all traffic; For all ports ,and Attributes are A collection of traffic types; express Attributes are Type of traffic on port Total flow rate The calculation formula is: ; express Attributes are Type of traffic on port Traffic bursts The calculation formula is: ; according to and get Attributes are Type of traffic on port arrival curve , The calculation formula is Where C represents the link bandwidth of the TSN network, and t represents the traffic transmission time; Step 203: Calculate the service curves at network nodes after CLP constraints; definition For traffic service rate, Initial service delay for the port; express Attributes are Type of traffic on port The traffic service rate is equal to the rate provided by the CBS mechanism. Credit growth rate of traffic type configuration ; express In the flow Attributes are Type of traffic on port Initial service delay; by Traffic service rate after parameter constraints The calculation formula is: Where C represents the link bandwidth of the TSN network; affected by Initial service delay of the port after parameter constraints The calculation formula is: ; according to and get Attributes are Type of traffic on port service curve , The calculation formula is , where t represents the traffic transmission time; Step 3: Calculate and verify the burstiness and latency guarantee of the traffic; Step 301: Based on the constrained arrival and service curves, calculate the maximum burstiness and maximum latency of traffic at a single node; (a) Maximum burstiness: Burst rate refers to the maximum amount of data that a queue can continuously send, defined as follows: ;by Port after parameter constraints middle Attributes are Burst of traffic type The calculation formula is ,in For flow through the port Maximum frame length; If any traffic of Attributes are When the type is used, the flow rate is high. At the port suddenness equal The calculation formula is: ,in For flow through the port Maximum frame length; (b) Maximum delay: flow any transmission path Recorded as , where nodes Marked as the current node, node It is a node The previous node, node It is a node The next node; Based on the flow arrival curves of the node output ports obtained in steps 202 and 203 and node service curve The port is calculated. middle Attributes are Maximum evaluation latency for this type of traffic The calculation formula is: ; In network calculus, the maximum horizontal intercept used to represent the arrival curve and the service curve is considered to be the maximum latency value. If any traffic of Attributes are When the type is used, the flow rate is high. At the port Maximum evaluation delay equal The calculation formula is: ; Step 302: Compare the calculated end-to-end latency with the requirements of the service flow to verify whether it meets the deterministic latency guarantee. For each traffic According to traffic transmission path All nodes that the transmission passes through suddenness Compare and take the larger value to get the flow rate. In the transmission path Maximum burst The calculation formula is: ; For each traffic According to traffic transmission path All nodes that the transmission passes through Maximum delay Sum them up to get the flow rate. In the transmission path End-to-end maximum delay The calculation formula is: ; Based on the calculated end-to-end delay of the traffic, determine any traffic... Does it meet the transmission objective of deterministic delay guarantee, that is, in terms of determinism, for any current traffic... The transmission time is less than the end-to-end transmission deadline of the traffic; Step 303: Iteratively adjust the CLP parameters and jump to step 201 until the optimal CLP value that minimizes traffic burstiness is found while satisfying the end-to-end delay constraint. Based on the calculation results of the deterministic delay guarantee in step 302, update Size; If the calculation result of step 302 meets the end-to-end delay requirement, then the time delay should be appropriately reduced in the next iteration. The parameter value is adjusted to achieve a smaller burst. If the calculation result of step 302 does not meet the end-to-end delay requirement, then the delay should be appropriately increased in the next iteration. The parameter value is used to satisfy any traffic. The latency requirements; If any flow condition is satisfied Under the premise of end-to-end latency requirements, The minimum value has been obtained, that is, the minimum value has been found. If the optimal value is found, then proceed to step 304; Step 304: Monitor the network operation status. If the traffic characteristics or network topology change, jump to step one and set the CLP parameters for the next round; otherwise, output the optimal CLP configuration value.

2. The parameter configuration method for burst control and delay guarantee in dynamic load networks as described in claim 1, characterized in that... The following steps are involved: the nodes in the network topology are connected by directed edges.

3. The parameter configuration method for burst control and delay guarantee in dynamic load networks as described in claim 1, characterized in that... The steps are as follows: By jumping to step 201, ensure that, under the premise of strictly satisfying all traffic deterministic latency guarantees, the optimal solution that minimizes traffic burstiness is successfully found. Configure values ​​to achieve the best balance between emergency control and delay protection.

4. The parameter configuration method for burst control and delay guarantee in dynamic load networks as described in claim 1, characterized in that... The process involves the following steps: In step 304, the network's operational status will be continuously monitored. If the traffic characteristics or network topology of the TSN network change, and the burstiness requirements and deterministic latency guarantees of the traffic in the TSN network cannot be met, the process will jump to step one to set the next round of CLP parameters, find the optimal CLP configuration value, and achieve the best balance between burstiness control and latency guarantee.

Images