Event-triggered traffic online admission control method for TSN network

Abstract

The invention discloses an event-triggered traffic online admission control method for a TSN network. The method comprises the following steps: S1, acquiring a network information set; s2, a request traffic information set is obtained, and the request traffic information set comprises a request traffic parameter tuple and request category information of request traffic; s3, when the request category information of the request traffic is an added traffic, processing the network information set and the request traffic information set by adopting a first traffic admission control model to complete admission control of the request traffic; and S4, when the request category information of the request flow is a deleted flow, processing the network information set and the request flow information set by using a second flow admission control model to complete admission control of the request flow.

Description

Technical Field

[0001] The present invention relates to the fields of industrial data processing, test evaluation, and data modeling processing, and particularly relates to an online admission control method for event-triggered traffic in a TSN network. Background Art

[0002] The rapid development of the industrial field has driven the demand for a deterministic communication network with dynamic reconfigurability. Time-Sensitive Networking (TSN) integrates Centralized Network Configuration (CNC) and Centralized User Configuration (CUC) through the IEEE 802.1Qcc standard to achieve dynamic configuration. However, although the 802.1Qcc standard specifies the interfaces and protocols, there are still many unresolved issues in the specific configuration methods. The main challenge of deterministic network reconfiguration lies in enhancing the utility of online admission control while ensuring the real-time requirements of incremental traffic and existing traffic, such as maximizing the number of admitted traffic and resource utilization, and shortening the reconfiguration time. For time-triggered (TT) traffic implemented through scheduling mechanisms such as Time-Aware Shaper (TAS) and Cyclic Queuing Forwarding (CQF), the configuration involves designing a scheduling table and requires combining deadline constraints to ensure the feasibility of scheduling. To reduce the computational complexity of static configuration methods, various online admission control methods for TT traffic in dynamic scenarios have been developed to support the dynamic addition and deletion of traffic. For event-triggered (ET) traffic implemented through scheduling mechanisms such as Strict Priority (SP), Credit-Based Shaper (CBS), and Asynchronous Traffic Shaper (ATS), the configuration only requires setting key parameters. Compared with TT traffic, ET traffic configuration is more flexible and is thus crucial for supporting real-time communication. However, different from TT traffic, ET traffic requires specialized performance analysis to ensure deadline satisfaction as feedback verification during configuration optimization. Research shows that performance analysis consumes more than 90% of the total configuration time, which is unacceptable in dynamic reconfiguration scenarios.

[0003] How to achieve online admission control for time-critical ET traffic in TSN is a problem that needs to be solved currently. Summary of the Invention

[0004] The present invention mainly solves the problem of how to realize online admission control of time-critical ET traffic in TSN, and discloses a method and device for online admission control of event-triggered traffic in a TSN network.

[0005] In a first aspect, an embodiment of the present invention discloses an event-triggered traffic online admission control method for a TSN network, comprising:

[0006] S1, obtain network information set;

[0007] S2, obtaining a request flow information set, wherein the request flow information set includes a request flow parameter tuple and request category information of the request flow;

[0008] S3, when the request category information of the request flow is to add a flow, the network information set and the request flow information set are processed using the first flow admission control model to complete the admission control of the request flow;

[0009] S4, when the request category information of the request flow is a delete flow, a second flow admission control model is used to process the network information set and the request flow information set to complete the admission control of the request flow.

[0010] The network information set satisfies a combination of constraint conditions; the expression of the combination of constraint conditions is:

[0011]

[0012]

[0013]

[0014]

[0015] Among them, (u1,v1)∈r f Indicates the route r f The output port (u1,v1) on Indicates the route r f The local deadline of the output port (u1,v1) on represents the set of traffic of the i-th type on the output port (u,v), l max Indicates the maximum frame length of all traffic in the network. Indicates the bandwidth of the j-th class of traffic on the output port (u, v), i and j are the serial numbers of the class information, represents the local deadline for the traffic of the i-th class on the output port (u,v).

[0016] Performing admission control on the request traffic by processing the network information set and the request traffic information set using the first traffic admission control model, including:

[0017] S31. Reading the candidate route set parameter set; the candidate route set parameter set includes candidate route set parameters; the candidate route set parameters are a set of candidate route sets pre-generated offline for each combination of source system to destination system, and the candidate route set contains the shortest k routes from the source system to the destination system; the candidate route set for the combination of source system s and destination system d is denoted as

[0018] S32. For the added traffic f + , according to its source system and destination system obtaining the corresponding candidate route set from the candidate route set parameter set

[0019] S33. For each + in the candidate route set parameter set corresponding to the added traffic f performing adjustment calculation of the local deadline to obtain the adjusted local deadline;

[0020] S34. Performing minimum bandwidth allocation calculation on the adjusted local deadline to obtain the adjusted minimum allocated bandwidth;

[0021] S35. Based on the adjusted minimum allocated bandwidth, performing feasibility analysis processing on the candidate route set to obtain the candidate route set Feasible;

[0022] S36. Performing optimal route planning processing on the candidate route set Feasible to obtain the optimal route information of the added traffic f + and completing the admission control of the request traffic.

[0023] The adjustment calculation of the local deadline includes:

[0024] Adjusting and calculating the local deadline of class i of the traffic belonging to the added traffic on the output port (u, v) of the route to obtain the adjusted local deadline The expression of the adjustment calculation is:

[0025]

[0026] where is the adjusted local deadline of the traffic of class i on the output port (u, v).

[0027] The local deadline adjustment model includes:

[0028] S331. For each output port (u, v) on the route, calculate the intermediate bandwidth and the remaining available bandwidth R , and the corresponding calculation expression is: (u,v)

[0029]

[0030] where {f +} represents the set of added traffic f + , and the initial local deadline of each strictly time-triggered AVB (Audio Video Bridging) traffic class j is and represent the intermediate bandwidth of the k-th type of traffic on the output port (u, v) and the intermediate bandwidth of the j-th type of traffic on the output port (u, v), represents the sum of all b in the subset of the i-th type of traffic on the output port (u, v) for the set f .

[0031] S332. For each output port (u, v) on the route, calculate the corresponding minimum possible local deadline; the calculation expression for the minimum possible local deadline is:

[0032]

[0033] where represents the minimum possible local deadline of the i-th type of traffic of the output port (u, v) on the route, {f } represents the set composed of the added traffic f + , + represents the intermediate bandwidth of the i-th type of traffic on the output port (u, v), and Φ i represents the additional bandwidth allocated to the i-th type of traffic.

[0034] S333. Make a feasibility judgment on the minimum possible local deadline to obtain a feasibility judgment result; if the feasibility judgment result is feasible, execute S334, and if the feasibility judgment result is infeasible, it indicates that there is no feasible local deadline, and end the admission control of the requested traffic;

[0035] S334. Initialize the first iteration variable, the second iteration variable, and the slack; let the first iteration variable γ = 1, the second iteration variable δ = 1, and the slack slack = -1;

[0036] S335, perform an updated calculation on the adjusted local deadline; the expression for the updated calculation is:

[0037]

[0038] where, is the adjusted local deadline of the traffic of the i-th type on the output port (u, v);

[0039] S336, perform parameter update processing on the slack, the first iteration variable, and the second iteration variable; the expression for the parameter update processing is:

[0040]

[0041] δ = δ / 2,

[0042]

[0043] S337, determine whether the slack satisfies 0 ≤ slack ≤ Threshold. If it is satisfied, confirm as the output result of the local deadline adjustment model; if it is not satisfied, execute S335.

[0044] The expression for the feasibility judgment is:

[0045]

[0046] If the above expression is satisfied, the feasibility judgment result is feasible; if the above expression is not satisfied, the feasibility judgment result is infeasible.

[0047] The expression for the minimum bandwidth allocation calculation is:

[0048]

[0049] where, is the adjusted minimum allocated bandwidth of the traffic of the i-th type on the output port (u, v), represents the sum of ρ for all traffic in the subset of the traffic of the i-th type on the output port (u, v). f Perform summation.

[0050] Perform feasibility analysis processing on the candidate route set to obtain the candidate route set Feasible, including:

[0051] Determine each candidate route in the candidate route set Whether the output result can be calculated in the local deadline adjustment model and the sum of the adjusted minimum allocated bandwidth of all AVB class traffic of the candidate route does not exceed the given upper limit idSl max ,Right now

[0052] From the candidate routing set All candidate routes that meet the above two conditions are screened out, and a candidate route set Feasible is constructed using all the screened candidate routes.

[0053] The expression of the optimal route planning process is:

[0054]

[0055] in, Candidate routes The cost function is To add flow f + The optimal routing information.

[0056] The adopting the second flow admission control model to process the network information set and the request flow information set to complete the admission control of the request flow includes:

[0057] S41, obtain deletion flow f - Routing information and the local deadline for each output port (u,v) on the route

[0058] S42, adjusting and updating the local deadline to obtain a deletion adjustment local deadline;

[0059] The expression for adjusting and updating the process is:

[0060]

[0061] in, Indicates the route The removal of traffic of class i on output port (u,v) adjusts the local deadline, Representing a collection Delete f - The collection constructed later;

[0062] S43, based on the deletion and adjustment of the local deadline, a minimum bandwidth allocation value that satisfies the deadline is calculated, and the calculation expression is:

[0063]

[0064] in, The minimum bandwidth allocation value for the event-triggered traffic class i that is strict for each time on each output port (u, v);

[0065] S44, based on the minimum bandwidth allocation value, perform bandwidth allocation for each output port of the routing information to complete the admission control for the deleted flow.

[0066] The beneficial effects of the present invention are as follows:

[0067] Based on the Network Calculus theory, the present invention proposes an online admission control method for time-critical ET traffic in TSN. This method is based on the TSN / ATS+CBS architecture, provides a routing selection and bandwidth allocation scheme for traffic addition requests, and provides a bandwidth recovery scheme for traffic deletion requests. On the basis of ensuring that time-critical ET traffic meets the end-to-end deadline, this method achieves fast response and high resource utilization. Description of the Drawings

[0068] Figure 1 is the implementation flowchart of the method of the present invention;

[0069] Figure 2 is the read network topology diagram of the third embodiment of the present invention. Detailed Embodiments

[0070] To better understand the content of the present invention, two embodiments are given here.

[0071] Embodiment 1:

[0072] Figure 1 is the implementation flowchart of the method of the present invention.

[0073] In the first aspect of the embodiment of the present invention, an online admission control method for event-triggered traffic in a TSN network is disclosed, including:

[0074] S1, obtaining a network information set;

[0075] S2, obtaining a request traffic information set, where the request traffic information set includes a request traffic parameter tuple and request category information of the request traffic;

[0076] S3, when the request category information of the request traffic is flow addition, using a first traffic admission control model to process the network information set and the request traffic information set to complete the admission control for the request traffic;

[0077] S4. When the request category information of the request traffic is the deletion flow, the second traffic admission control model is used to process the network information set and the request traffic information set to complete the admission control of the request traffic.

[0078] The network information set includes network topology information, initial configuration parameter information, and initial traffic parameter set information;

[0079] The network topology information is expressed as where is a set of nodes, and the nodes include end systems and switches. The end systems include source end systems and destination end systems, is a set of physical links. represents the physical link from node u to node v and the corresponding output port, that is, the output port of node v. The rate of this physical link is C. There are only nodes u to node v on the physical link of (u, v), and no other nodes. The source end system and the destination end system are the starting node and the destination node of a traffic respectively.

[0080] The initial configuration parameter information includes the initial local deadline of the AVB (Audio Video Bridging) traffic class i triggered by strict events for each time on each output port (u, v) in the network and the initial bandwidth the upper bound idSL of the bandwidth that can be allocated to all AVB traffic on all output ports max and the number N of AVB classes AVB .

[0081] The initial traffic parameter set information includes the parameter tuple of each traffic f, the class information to which the traffic f belongs, the route r to which the traffic f belongs f and on the route r f the local deadline of each output port (u, v) the maximum frame length l of all traffic in the network max and the maximum frame length l of the Best Effort (BE) traffic BE , the maximum burst data volume per unit time of each traffic and the traffic rate;

[0082] The parameter tuple of the traffic f is expressed as where s f is the source end system of the traffic f, d f is the destination end system of the traffic f, l f is the frame length of the traffic f, p f is the period of the traffic f, is the end-to-end deadline of the traffic f; for each traffic f, b fis the maximum burst data volume per unit time, b f = l f , the rate ρ of the traffic f f = l f / p f .

[0083] For the requested traffic parameter tuple for adding traffic f + , including the parameter tuple expression, the maximum burst data volume per unit time, and the rate;

[0084] The parameter tuple expression for adding traffic f + is is the source system of traffic f + , is the destination system of traffic f + , is the frame length of traffic f + , is the period of traffic f + , is the end-to-end deadline of traffic f + ; The maximum burst data volume per unit time for adding traffic f + is Its rate is The class it belongs to is i;

[0085] For the requested traffic parameter tuple for deleting traffic f - , including the parameter tuple expression, the maximum burst data volume per unit time, and the rate;

[0086] The parameter tuple expression for deleting traffic f - is is the source system of traffic f - , is the destination system of traffic f - , is the frame length of traffic f - , is the period of traffic f - , is the end-to-end deadline of traffic f - ; The maximum burst data volume per unit time for deleting traffic f - is Its rate is The class it belongs to is i;

[0087] The network information set satisfies the combination of constraint conditions; The expression of the combination of constraint conditions is:

[0088]

[0089]

[0090]

[0091]

[0092] where (u1, v1) ∈ r f represents the output port (u1, v1) on the routing r f and represents the local deadline of the output port (u1, v1) on the routing r f ; represents the set of traffic of the i-th class on the output port (u, v), l max represents the maximum frame length of all traffic in the network represents the bandwidth of the traffic of the j-th class on the output port (u, v), where i and j are the sequence numbers of class information represents the local deadline of the traffic of the i-th class on the output port (u, v);

[0093] The first traffic admission control model is used to process the network information set and the requested traffic information set to complete the admission control of the requested traffic, including:

[0094] S31. Read the candidate routing set parameter set; the candidate routing set parameter set includes candidate routing set parameters; the candidate routing set parameters are a set of candidate routing sets pre-generated offline for each combination of source system to destination system, and the candidate routing set contains the shortest k routes from the source system to the destination system; the candidate routing set for the combination of source system s and destination system d is denoted as

[0095] S32. For the added traffic f + , according to its source system and destination system , obtain the corresponding candidate routing set from the candidate routing set parameter set

[0096] S33. For each candidate routing set parameter set corresponding to the added traffic f + , perform adjustment calculation of the local deadline to obtain the adjusted local deadline; ;

[0097] S34. Perform minimum bandwidth allocation calculation on the adjusted local deadline to obtain the adjusted minimum allocated bandwidth;

[0098] S35. Based on the adjusted minimum allocated bandwidth, for the candidate routing set Perform a feasibility analysis process to obtain a candidate route set Feasible;

[0099] S36. Perform an optimal route planning process on the candidate route set Feasible to obtain the added traffic f + of the optimal route information, and complete the admission control of the requested traffic.

[0100] The adjustment calculation of the local deadline includes:

[0101] For a route when, in order to meet its end-to-end deadline For a route the local deadline for class i of the added traffic on the output port (u, v) on is adjusted and calculated to obtain the adjusted local deadline The expression of the adjustment calculation is:

[0102]

[0103] Where, is the adjusted local deadline for the traffic of class i on the output port (u, v).

[0104] The local deadline adjustment model includes:

[0105] S331. For each output port (u, v) on a route calculate the intermediate bandwidth and the remaining available bandwidth R (u,v) , and the corresponding calculation expression is:

[0106]

[0107] Where, {f +} represents the set of added traffic f + The initial local deadline for each strictly time-triggered AVB (Audio Video Bridging) traffic class j is and represent the intermediate bandwidth of the traffic of class k on the output port (u, v) and the intermediate bandwidth of the traffic of class j on the output port (u, v), represents, for the set all the b of the traffic within the subset of the traffic of class i on the output port (u, v) f are summed up.

[0108] S332. For a route For each output port (u, v) on it, calculate the corresponding minimum possible local deadline; the calculation expression of the minimum possible local deadline is:

[0109]

[0110] where represents the minimum possible local deadline of the traffic of the i-th class on the output port (u, v) of the route {f +} represents the set formed by adding the traffic f + , represents the intermediate bandwidth of the traffic of the i-th class on the output port (u, v), and Φ i represents the additional bandwidth allocated to the traffic of the i-th class.

[0111] S333. Make a feasibility judgment on the minimum possible local deadline to obtain a feasibility judgment result; if the feasibility judgment result is feasible, execute S334; if the feasibility judgment result is infeasible, it indicates that there is no feasible local deadline, and end the admission control of the requested traffic;

[0112] S334. Initialize the first iteration variable, the second iteration variable, and the slack. Let the first iteration variable γ = 1, the second iteration variable δ = 1, and the slack slack = -1;

[0113] S335. Update and calculate the adjusted local deadline; the expression of the update calculation is:

[0114]

[0115] where is the adjusted local deadline of the traffic of the i-th class on the output port (u, v);

[0116] S336. Perform parameter update processing on the slack, the first iteration variable, and the second iteration variable; the expression of the parameter update processing is:

[0117]

[0118] δ = δ / 2,

[0119]

[0120] S337. Judge whether the slack satisfies 0 ≤ slack ≤ Threshold. If it satisfies, confirm as the output result of the local deadline adjustment model; if it does not satisfy, execute S335; Threhold is a preset slack discrimination threshold.

[0121] The expression of the feasibility judgment is:

[0122]

[0123] If the above expression is satisfied, the feasibility judgment result is feasible; if the above expression is not satisfied, the feasibility judgment result is infeasible.

[0124] The expression for calculating the minimum bandwidth allocation is:

[0125]

[0126] in, is the adjusted minimum allocated bandwidth for the i-th class of traffic on the output port (u,v), Indicates that for the set ρ of all flows in the subset of flows of type i at output port (u,v) f Perform the summation.

[0127] The candidate routing set Perform feasibility analysis to obtain a set of candidate routes, Feasible, including:

[0128] Determine the candidate routing set Each candidate route in Whether the output result can be calculated in the local deadline adjustment model and the sum of the adjusted minimum allocated bandwidth of all AVB class traffic of the candidate route does not exceed the given upper limit idSl max ,Right now

[0129] From the candidate routing set All candidate routes that meet the above two conditions are screened out, and a candidate route set Feasible is constructed using all the screened candidate routes.

[0130] The expression of the optimal route planning process is:

[0131]

[0132]

[0133] in, Candidate routes The cost function is To add flow f + The optimal routing information.

[0134] Performing admission control on the request traffic by processing the network information set and the request traffic information set using the second traffic admission control model, including:

[0135] S41. Obtain the routing information of the deletion traffic f - and the local deadline of each output port (u, v) on the route

[0136] S42. Perform adjustment and update processing on the local deadline to obtain the adjusted local deadline for deletion;

[0137] The expression of the adjustment and update processing is:

[0138]

[0139] where represents the adjusted local deadline for deletion of the traffic of the i-th class on the output port (u, v) of the route represents the set constructed after deleting f

[0140] from the set - ;

[0141] S43. Based on the adjusted local deadline for deletion, calculate the minimum bandwidth allocation value that meets the deadline, and its calculation expression is:

[0142] where

[0143] is the minimum bandwidth allocation value for the traffic class i triggered by events that are strict with respect to time on each output port (u, v); S44. Based on the minimum bandwidth allocation value, allocate the bandwidth of each output port of the routing information

[0144] to complete the admission control of the deletion flow.

[0145] The traffic class i is the traffic class of AVB. The Φ i is obtained from the starting term of and the recurrence relation of

[0146] , and the recurrence relation is

[0147]

[0148] where

[0149] ​​

[0150]

[0151] Among them, η, ξ, and ζ are the first intermediate quantity, the second intermediate quantity, and the third intermediate quantity respectively, and Φ k represents the additional bandwidth of the traffic allocated to the k-th class; for the starting term Φ1, by taking i = 1 and j = 2, or by backward derivation from the high term to the low term, and g() is a recurrence function.

[0152] For the offline pre-generation, the network can be represented as a directed graph, and the shortest path planning algorithm is used to solve for each source system to the destination system to obtain a set of candidate routing sets.

[0153] Embodiment 2:

[0154] An embodiment of the present invention discloses an event-triggered traffic online admission control method for a TSN network, including:

[0155] Step 1: Read the network topology, initial configuration, initial traffic, requested traffic, and candidate routing set. Step 1-1: Read the network topology and initial configuration parameters

[0156] Read the network topology and initial configuration parameters through an input file. Read the network topology parameters where is the set of nodes (including end systems and switches), is the set of physical links. represents both the physical link from node u to node v and the corresponding output port, and its rate is C. Read the initial local deadline and the initial bandwidth of the event-triggered AVB (Audio Video Bridging) traffic class i for each time-strict on each output port (u, v) in the network. max In addition, read the bandwidth upper bound idSl AVB that the output port can allocate to all AVB traffic and the number N

[0157] of AVB classes.

[0158] Step 1-2: Read the initial traffic parameters Read the initial traffic parameters through an input file. Read the parameter tuple of each traffic f in the initial traffic set f in the network, where s f is the source end system, d fis the frame length, p f is the period, is the end - to - end deadline. And for each flow f, its burst b f = l f , its rate ρ f = l f / p f . In addition, it is also necessary to read the class to which the flow f belongs, the route r f and the local deadline at each output port on the route In addition, read the maximum frame length l of the flows in the network max and the maximum frame length l of the BE flows BE .

[0159] Step 1 - 3: Read the requested traffic parameters

[0160] Read the requested traffic parameters through the input file. For the flow addition request, read the new flow f + whose parameter tuple whose burst is whose rate is and the class to which it belongs is i; for the flow deletion request, read the deleted flow f - whose parameter tuple burst rate route and the local deadline at each output port on the route and the class to which it belongs is i.

[0161] Step 1 - 4: Read the candidate route set parameters

[0162] Read the candidate route set parameters through the input file. To avoid the complexity of online routing, we pre - generate a set of candidate route sets for each source - destination pair offline. Each candidate route set contains the k shortest routes from the source node to the destination node for efficient online retrieval. The candidate route set for the source - destination pair of end - system s and end - system d is denoted as

[0163] Step Two: Retrieve the candidate route set

[0164] For the newly added flow f for the request + , we allocate the corresponding candidate route set according to its source node and destination node and the candidate route set group obtained in Step 1 - 4

[0165] Step Three: Calculate the minimum bandwidth allocation with deadline adaptation

[0166] Step 3-1: Calculate the adjusted local deadline

[0167] When the new flow f + is assigned to the route in order to meet its end-to-end deadline it is necessary to adjust the local deadline for class i to which the new flow belongs on the output port (u, v) of this route . The adjusted local deadline is

[0168]

[0169] where the specific steps of Algorithm 1 are as follows. For the network output port (u, v) outside the route , the local deadline remains unchanged, that is, the adjusted local deadline is Specific steps of Algorithm 1:

[0170] 1. Calculate the remaining available bandwidth

[0171] For each output port (u, v) on the route , calculate the remaining available bandwidth r (u,v) as

[0172]

[0173] where

[0174]

[0175] 2. Calculate the minimum possible local deadline

[0176] For each output port (u, v) on the route , calculate the minimum possible local deadline as

[0177]

[0178] where Φ i is obtained from the starting term of and the recurrence relation of , and the recurrence relation is

[0179]

[0180] where

[0181]

[0182] 3. Feasibility determination

[0183] If the following conditions are met:

[0184]

[0185] Then proceed to the next calculation Otherwise, it indicates that there is no feasible local deadline.

[0186] 4. Optimize and adjust the local deadline

[0187] Initialization: Let γ = 1, δ = 1, slack = -1.

[0188] Iterative calculation:

[0189] For each output port (u, v) on the route calculate as

[0190]

[0191] where Φ i is obtained from the starting term of and the recurrence relation of

[0192] Calculate the slack as

[0193]

[0194] Update γ and δ to be respectively

[0195] δ = δ / 2,

[0196]

[0197] Repeat the above steps until 0 ≤ slack ≤ Threshold is satisfied, and obtain

[0198] Step 3 - 2: Calculate the adjusted minimum bandwidth allocation

[0199] According to the adjusted deadline obtained in Step 3 - 1 calculate the adjusted minimum bandwidth allocation scheme as

[0200]

[0201] Step Four: Select the optimal route and bandwidth allocation scheme for the added flow

[0202] Step 4 - 1: Conduct feasibility analysis on each candidate route

[0203] For each candidate route conduct feasibility analysis on it and the corresponding bandwidth allocation scheme. A candidate route To be considered feasible, the following two conditions must be met simultaneously: (1) Step 3-1 can obtain a feasible local deadline; (2) On each output port (u, v) in the network, the sum of the bandwidths of all AVB class allocations does not exceed the given upper limit idSl max , that is

[0204] Step 4-2: Optimal configuration selection

[0205] If there is at least one feasible route, we select the optimal route by the following method and admit the new flow f using this optimal route and the corresponding bandwidth allocation scheme + . If there is no feasible route, a route that meets the end-to-end deadline of the new flow cannot be found, and the request for the new flow will be rejected.

[0206] Optimal route is defined as follows:

[0207]

[0208] Among them, the cost function of the route is

[0209]

[0210] Step Five: Provide a bandwidth recovery scheme for the deleted flow

[0211] When the request is to remove the flow f - , remove the local deadline related to the flow f on the output port (u, v) of the removal route For the output port (u, v) on the route - , the adjusted local deadline For the route on the output port (u, v), the adjusted local deadline is

[0212]

[0213] For the output port (u, v) outside the route , the local deadline remains unchanged, that is

[0214] According to the obtained adjusted deadline calculate the minimum bandwidth allocation scheme that meets this deadline That is, the bandwidth allocation scheme after removing the flow f - is

[0215]

[0216] Example Three:

[0217] An embodiment of the present invention discloses an online admission control method for event-triggered traffic in a TSN network, including:

[0218] Case 1: Add flow request

[0219] Step 1: Read the network topology, initial configuration, initial traffic, request traffic, and candidate route set. Step 1-1: Read the network topology and initial configuration parameters

[0220] The read network topology is as Figure 2 shown, and the specific topology parameter information is shown in Table 1. In addition, the upper bound of the bandwidth idSl max allocated to all AVB traffic is 0.75C, and the number N AVB of AVB classes is 2.

[0221] Table 1 Network initial configuration parameters

[0222]

[0223] Step 1-2: Read the initial traffic parameters

[0224] The read initial traffic parameters are shown in Table 2. In addition, the maximum frame length l max of the traffic in the network is 1518 Byte, and the maximum frame length l BE of the BE traffic in the network is 1518 Byte. The network initial configuration parameters in Table 1 and the network initial traffic parameters in Table 2 satisfy the relationship given in Step 1-2 of the specific implementation method.

[0225] Table 2 Network initial traffic parameters

[0226]

[0227]

[0228] Step 1-3: Read the request traffic parameters

[0229] The read request is an add flow request, and the parameters of the newly added flow f5 are shown in Table 3.

[0230] Table 3 Request traffic parameters

[0231]

[0232] Step 1-4: Read the candidate route set parameters

[0233] Select the parameter k = 2, that is, select the two shortest paths. The read candidate route set parameters are shown in Table 4.

[0234] Table 4 Candidate route set

[0235]

[0236] Step 2: Retrieve the candidate route set

[0237] Step 2-1: Provide the candidate route set

[0238] For the new flow f5, its source node is ES1 and its destination node is ES3. According to Table 4 in Step 1-4, its candidate routes are Route 1: (ES1, SW1), (SW1, SW3), (SW3, ES3) and Route 2: (ES1, SW1), (SW1, SW2), (SW2, SW3), (SW3, ES3).

[0239] Step 3: Calculate the deadline-adaptive minimum bandwidth allocation

[0240] Step 3-1: Calculate the adjusted local deadline

[0241] Route 1: (ES1, SW1), (SW1, SW3), (SW3, ES3)

[0242] The sum of the local deadlines on Route 1 is 12000 us, which is less than or equal to the end-to-end deadline of the new flow f5, which is 12000 us, that is Therefore, the local deadline of the output port on Route 1 is adjusted to For the output ports other than those on Route 1, the adjusted local deadline is When using Route 1, the adjusted local deadlines are shown in Table 5.

[0243] Table 5 Adjusted local deadlines (using Route 1)

[0244]

[0245] Route 2: (ES1, SW1), (SW1, SW2), (SW2, SW3), (SW3, ES3). The sum of the local deadlines on Route 2 is 16000 us, which is greater than the end-to-end deadline of the new flow f5, which is 12000 us, that is Therefore, the adjusted local deadlines of the output ports on Route 2 are obtained through Algorithm 1. For the output ports other than those on Route 2, the adjusted local deadline is The specific steps of Algorithm 1 are as follows:

[0246] 1. Calculate the remaining available bandwidth

[0247] Table 6 Calculate the remaining available bandwidth

[0248]

[0249] 2. Calculate the minimum possible local deadline

[0250] Table 7 Calculate the minimum possible local deadline

[0251]

[0252] 3. Feasibility determination

[0253] Meet the conditions Proceed to the next calculation

[0254] 4. Optimize and adjust the local deadline

[0255] Select the parameter Threshold = 1us, and the optimized and adjusted local deadline is shown in Table 8

[0256] Table 8 Optimize and adjust the local deadline

[0257]

[0258] According to the adjusted local deadline of the output port on Route 2 in Table 8, the adjusted local deadline using Route 2 is shown in Table 9

[0259] Table 9 Adjusted local deadline (using Route 2)

[0260]

[0261] Step 3-2: Calculate the adjusted minimum bandwidth allocation

[0262] Route 1: (ES1, SW1), (SW1, SW3), (SW3, ES3)

[0263] Table 10 Adjusted minimum bandwidth allocation (using Route 1)

[0264]

[0265]

[0266] Route 2: (ES1, SW1), (SW1, SW2), (SW2, SW3), (SW3, ES3) Table 11 Adjusted minimum bandwidth allocation (using Route 2)

[0267]

[0268] Step Four: Select the optimal routing and bandwidth allocation scheme for adding flows

[0269] Step 4-1: Conduct a feasibility analysis for each candidate route

[0270] After step 3-1, Routing 1 and Routing 2 respectively obtain feasible local deadlines. And on each output port (u, v) in the network, the sum of the bandwidths allocated for all AVB classes does not exceed idSl max . Therefore, both Routing 1 and Routing 2 are feasible.

[0271] Step 4-2: Optimal configuration selection

[0272] Table 12 Cost function

[0273]

[0274] The cost functions of Routing 1 and Routing 2 are shown in Table 12. Since the cost function of Routing 1 is less than that of Routing 2, Routing 1 and the bandwidth allocation in Table 10 are selected as the optimal configuration for the admission of the new flow f5.

[0275] Case 2: Flow deletion request

[0276] Step 1: Read the network topology, initial configuration, initial traffic, request traffic, and candidate routing set Step 1-1: Read the network topology and configuration parameters

[0277] The read network topology is as Figure 2 shown, and the specific topology parameter information is shown in Table 13. In addition, the upper bound of the bandwidth allocated to all AVB traffic idSl max = 0.75C, and the number of AVB classes N AVB = 2.

[0278] Table 13 Network initial configuration parameters

[0279]

[0280] Step 1-2: Read the initial traffic parameters

[0281] The read initial traffic parameters are shown in Table 14. In addition, the maximum frame length l of the traffic in the network max is 1518 Byte, and the maximum frame length l of the BE traffic in the network BE is 1518 Byte. The network initial configuration parameters in Table 13 and the network initial traffic parameters in Table 14 satisfy the relationships given in step 1-2 of the specific implementation method.

[0282] Table 14 Network initial traffic parameters

[0283]

[0284]

[0285] Step 1-3: Read the request traffic parameters

[0286] The read request is a flow deletion request, and the parameters for deleting flow f4 are shown in Table 15.

[0287] Table 15 Request traffic parameters

[0288]

[0289] Step Five: Provide a bandwidth recovery solution for flow deletion

[0290] Step 5-1: Calculate the adjusted local deadline

[0291] The adjusted local deadline after deleting flow f4 is shown in Table 17.

[0292] Table 17 Adjusted local deadline

[0293]

[0294]

[0295] Step 5-2: Calculate the adjusted minimum bandwidth allocation

[0296] The adjusted minimum bandwidth allocation after deleting flow f4 is shown in Table 18.

[0297] Table 18 Adjusted minimum bandwidth allocation

[0298]

[0299] The above are only embodiments of the present invention and are not used to limit the present invention. For those skilled in the art, the present invention can have various changes and modifications. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included within the scope of the claims of the present invention.

Claims

1. A method for online admission control of event-triggered traffic in a TSN network, characterized in that: include: S1, obtain network information set; S2, obtaining a request flow information set, wherein the request flow information set includes a request flow parameter tuple and request category information of the request flow; S3, when the request category information of the request flow is to add a flow, the network information set and the request flow information set are processed using the first flow admission control model to complete the admission control of the request flow; S4, when the request category information of the request flow is a delete flow, a second flow admission control model is used to process the network information set and the request flow information set to complete the admission control of the request flow.

2. The method for online admission control of event-triggered traffic in a TSN network according to claim 1, characterized in that: The network information set includes network topology information, initial configuration parameter information and initial traffic parameter set information; the network topology information is represented as in is a collection of nodes, including end systems and switches, and end systems include source end systems and destination end systems. is a collection of physical links; represents the physical link from node u to node v and the corresponding output port, and the rate of the physical link is C; The initial configuration parameter information includes the initial local deadline for each time-critical event-triggered AVB traffic class i on each output port (u, v) in the network and initial bandwidth The upper bound of the bandwidth that can be allocated to all AVB traffic on all output ports is idSl max and the number of AVB classes N AVB ; The initial flow parameter set information includes the parameter tuple of each flow f, the class information to which the flow f belongs, and the route r to which the flow f belongs. f and in the routing f The local deadline for each output port (u,v) on The maximum frame length of all traffic in the network max and the maximum frame length l for Best Effort (BE) traffic BE , the maximum burst data volume and traffic rate per unit time for each flow; The parameter tuple of the flow f is expressed as where s f is the source system of flow f, d f is the destination system of flow f, l f is the frame length of flow f, p f is the period of flow f, is the end-to-end deadline of flow f; For each flow f, b f is the maximum burst data volume per unit time, b f = l f , the rate ρ of the flow f f = l f / p f ; For adding flow f + The request flow parameter tuple includes the parameter tuple expression, the maximum burst data volume and rate per unit time; add flow f + The parameter tuple expression is is the flow rate f + The source system, is the flow rate f + The destination system, is the flow rate f + The frame length, is the flow rate f + The cycle, is the flow rate f + end-to-end deadline; add flow f + The maximum burst data volume per unit time is Its rate is The class it belongs to is i; For deleting flow f - The request flow parameter tuple includes the parameter tuple expression, the maximum burst data volume and rate per unit time; delete the flow f - The parameter tuple expression is is the flow rate f - The source system, is the flow rate f - The destination system, is the flow rate f - The frame length, is the flow rate f - The cycle, is the flow rate f - end-to-end deadline; remove flow f - The maximum burst data volume per unit time is Its rate is The class it belongs to is i.

3. The event-triggered traffic online admission control method of the TSN network as claimed in claim 2, characterized in that: The network information set satisfies a combination of constraint conditions; the expression of the combination of constraint conditions is: Among them, (u1,v1)∈r f Indicates the route r f The output port (u1,v1) on Indicates the route r f The local deadline of the output port (u1,v1) on represents the set of traffic of the i-th type on the output port (u,v), l max Indicates the maximum frame length of all traffic in the network. Indicates the bandwidth of the j-th class of traffic on the output port (u, v), i and j are the serial numbers of the class information, represents the local deadline for the traffic of the i-th class on the output port (u,v).

4. The method for online admission control of event-triggered traffic in a TSN network as claimed in claim 2, characterized in that: The adopting the first flow admission control model to process the network information set and the request flow information set to complete the admission control of the request flow includes: S31, read a candidate route set parameter set; the candidate route set parameter set includes candidate route set parameters; the candidate route set parameters are a set of candidate route sets pre-generated offline for each combination of the source system to the destination system, and the candidate route set includes the shortest k routes from the source system to the destination system; the candidate route set for the combination of the source system s and the destination system d is expressed as S32, add flow f + , according to its source system and the destination system Obtain the corresponding candidate route set from the candidate route set parameter set S33, for adding flow f + Each of the corresponding candidate routing set parameters Performing an adjustment calculation of the local deadline to obtain an adjusted local deadline; S34, performing minimum bandwidth allocation calculation on the adjusted local deadline to obtain an adjusted minimum allocated bandwidth; S35, based on the adjusted minimum allocated bandwidth, Perform feasibility analysis and obtain a set of feasible candidate routes; S36, performing optimal route planning processing on the candidate route set Feasible to obtain the added traffic f + The optimal routing information is used to complete the admission control of the request traffic.

5. The method for online admission control of event-triggered traffic in a TSN network as claimed in claim 4, characterized in that: The adjustment calculation for the local deadline includes: For routing The local deadline on the output port (u,v) for adding traffic to class i Perform adjustment calculations to obtain the adjusted local deadline The expression of the adjustment calculation is: in, is the adjusted local deadline for traffic of class i at output port (u,v).

6. The method for online admission control of event-triggered traffic in a TSN network as claimed in claim 5, characterized in that: The local deadline adjustment model includes: S331, for routing For each output port (u,v) on the (u,v) , the corresponding calculation expression is: Among them, {f + } indicates adding flow f + The initial local deadline for each time-strict event-triggered AVB (Audio Video Bridging) traffic class j is and represents the median bandwidth of the k-th class of traffic on the output port (u, v) and the median bandwidth of the j-th class of traffic on the output port (u, v), Indicates that for the set b of all flows in the subset of flows of type i at output port (u,v) f Perform summation; S332, for routing For each output port (u, v) on , the corresponding minimum possible local deadline is calculated; the calculation expression of the minimum possible local deadline is: in, Indicates the route The minimum possible local deadline for the i-th type of traffic at the output port (u,v) on + } indicates adding flow f + The set composed of represents the median bandwidth of the i-th type of traffic on the output port (u,v), Φ i represents the additional bandwidth allocated to the i-th class of traffic; S333, performing feasibility judgment on the minimum possible local deadline to obtain a feasibility judgment result; if the feasibility judgment result is feasible, executing S334; if the feasibility judgment result is infeasible, indicating that there is no feasible local deadline, the admission control of the request traffic is terminated; S334, initializing the first iteration variable, the second iteration variable and the slack, setting the first iteration variable γ=1, the second iteration variable δ=1, and the slack slack=-1; S335, performing update calculation on the adjusted local deadline; the expression of the update calculation is: in, is the adjusted local deadline for the flow of the i-th class on the output port (u,v); S336, performing parameter update processing on the slack amount, the first iteration variable and the second iteration variable; the expression of the parameter update processing is: δ=δ / 2, S337, determine whether the slack satisfies 0≤slack≤Threshold, if so, confirm The output result of the model is adjusted for the local deadline; if not satisfied, executing S335.

7. The method for online admission control of event-triggered traffic in a TSN network as claimed in claim 6, characterized in that: The expression of the feasibility judgment is: If the above expression is satisfied, the feasibility judgment result is feasible; if the above expression is not satisfied, the feasibility judgment result is infeasible.

8. The method for online admission control of event-triggered traffic in a TSN network as claimed in claim 4, characterized in that: The expression for calculating the minimum bandwidth allocation is: in, is the adjusted minimum allocated bandwidth for the i-th class of traffic on the output port (u,v), Indicates that for the set ρ of all flows in the subset of flows of type i at output port (u,v) f Perform the summation.

9. The method for online admission control of event-triggered traffic in a TSN network as claimed in claim 4, characterized in that: The candidate routing set Perform feasibility analysis to obtain a set of candidate routes, Feasible, including: Determine the candidate routing set Each candidate route in Whether the output result can be calculated in the local deadline adjustment model and the sum of the adjusted minimum allocated bandwidth of all AVB class traffic of the candidate route does not exceed the given upper limit idSl max ,Right now From the candidate routing set All candidate routes that meet the above two conditions are screened out, and a candidate route set Feasible is constructed using all the screened candidate routes; The expression of the optimal route planning process is: in, The candidate route r f+ The cost function is To add flow f + The optimal routing information.

10. The method for online admission control of event-triggered traffic in a TSN network as claimed in claim 3, characterized in that: The adopting the second flow admission control model to process the network information set and the request flow information set to complete the admission control of the request flow includes: S41, obtain deletion flow f - Routing information and the local deadline for each output port (u,v) on the route S42, adjusting and updating the local deadline to obtain a deletion adjustment local deadline; The expression for adjusting and updating the process is: in, Indicates the route The removal of traffic of class i on output port (u,v) adjusts the local deadline, Representing a collection Delete f - The collection constructed later; S43, based on the deletion and adjustment of the local deadline, a minimum bandwidth allocation value that satisfies the deadline is calculated, and the calculation expression is: in, is the minimum bandwidth allocation value for traffic class i triggered by each time-strict event on each output port (u,v); S44, based on the minimum bandwidth allocation value, routing information The bandwidth of each output port is allocated to complete the admission control of the deletion flow.

Images