Bus network time-triggered traffic admission planning method giving consideration to rate-limited traffic

Abstract

The invention relates to a bus network time-triggered traffic admission planning method giving consideration to rate-limited traffic, and belongs to the technical field of wired communication network resource dynamic scheduling and planning. The method comprises the following steps: acquiring an optional routing scheme based on an access request of a new TT flow set; when a first routing scheme meeting the condition that the end-to-end worst condition time delay of each RC flow does not exceed the corresponding end-to-end time delay period exists in the selectable routing schemes, and a first scheduling scheme meeting the scheduling constraint exists based on the first routing scheme, executing the first scheduling scheme; and otherwise, selecting a second routing scheme in which the number of RC streams is as small as possible, the end-to-end worst case time delay of which exceeds the corresponding end-to-end time delay period, from the selectable routing schemes, and executing the second scheduling scheme when a second scheduling scheme meeting the scheduling constraint exists based on the second routing scheme. The method provided by the invention greatly reduces the probability of RC flow end-to-end delay timeout and even communication interruption during new TT flow admission planning in the prior art.

Description

Technical Field

[0001] The present invention belongs to the technical field of dynamic scheduling and planning of wired communication network resources, and in particular relates to a bus network time-triggered flow admission planning method taking into account rate-limited flow. Background Art

[0002] As the trend of merging Ethernet with bus networks of various integrated electronic systems such as aerospace, aviation, and vehicles becomes stronger and stronger, software-defined networks are increasingly introduced into real-time control systems, further enhancing the network's explicit control over traffic and promoting research on dynamic planning and configuration of time-triggered network traffic that previously required static verification and deployment.

[0003] At present, the research on dynamic programming of time-triggered networks (such as time-sensitive networks TSN) with multi-priority mixed criticality traffic mainly focuses on time-triggered (TT) traffic, including responding to network component failures (such as links), switching network application migration decisions, end-to-end delay performance optimization, or incremental design of TT flows. These studies explore dynamic programming methods for TT flows from the perspective of many applications, which can improve the reusability, reliability, and performance of TSN networks.

[0004] However, even if the scheduling verification is passed, the reconfiguration scheme of the TT flow (such as the addition of new traffic, etc.) may cause changes in the switch port load and network bandwidth resources, thereby affecting the end-to-end delay of rate-limited low-priority traffic (RC flow), and even causing them to miss the worst-case deadline, which means that the service quality of some relatively critical RC flows is reduced or even the communication continuity is destroyed, resulting in resource waste or more serious consequences. Summary of the invention

[0005] In view of the above analysis, the present invention aims to provide a bus network time-triggered traffic admission planning method that takes into account rate-limited traffic, minimizes the impact of dynamic planning of new TT flows to be admitted in the TSN network on RC flows, and avoids the problem of new TT flows affecting the end-to-end delay of RC flows as much as possible.

[0006] The present invention provides a bus network time-triggered traffic admission planning method taking into account rate-limited traffic, which specifically includes the following steps:

[0007] Based on the admission request of the received new TT flow set, multiple optional routing solutions are obtained one by one;

[0008] When there is a first routing solution in the optional routing solutions that satisfies the requirement that the end-to-end worst-case delay of each RC flow does not exceed the corresponding end-to-end delay limit, and there is a first scheduling solution based on the first routing solution that satisfies the scheduling constraint, the first scheduling solution is executed; otherwise,

[0009] A second routing scheme in which the number of RC flows whose end-to-end worst-case delay exceeds the corresponding end-to-end delay limit is as small as possible is selected from the optional routing schemes, and when a second scheduling scheme that satisfies the scheduling constraint exists based on the second routing scheme, the second scheduling scheme is executed.

[0010] Further, determining whether there is a first routing solution among the optional routing solutions that satisfies the requirement that the end-to-end worst-case delay of each RC flow does not exceed the corresponding end-to-end delay limit includes:

[0011] Iteratively solving an optimal routing solution based on the optional routing solution with the goal of optimizing the delay time rate of each RC flow affected by the new TT flow;

[0012] In each round of iteration, based on the optimal routing solution obtained in this round of iteration, it is determined whether the worst-case end-to-end delay of each RC flow does not exceed the corresponding end-to-end delay limit:

[0013] If yes, the optimal routing solution obtained in this round of iteration is the first routing solution;

[0014] If not, continue with the next iteration until the maximum number of iterations is reached.

[0015] Further, a particle swarm algorithm is used to optimize the delay time rate of each RC flow affected by the new TT flow, and an optimal routing solution is iteratively solved based on the optional routing solution, including:

[0016] Numbering the optional routing solutions;

[0017] A combination of the routing schemes of the new TT flows is used as a particle, a combination of corresponding numbers of the routing schemes of the new TT flows in each particle is used to describe the particle position, and a particle group is randomly initialized based on the multiple optional routing schemes;

[0018] A fitness function is constructed based on the delay time rate and maximum bandwidth constraint of the affected RC flows;

[0019] In each round of iteration, the optimal solution of the particle swarm is sought based on the fitness function to obtain the optimal routing solution for that round of iteration.

[0020] Further, constructing a fitness function based on the delay time rate and maximum bandwidth constraint of the affected RC stream includes:

[0021] An objective function is constructed based on the delay time rate of the affected RC flows;

[0022] Construct a penalty function based on the maximum bandwidth constraint;

[0023] The fitness function is constructed based on the objective function, the penalty function and the weight coefficient of the penalty function.

[0024] Furthermore, the objective function is expressed as:

[0025]

[0026] Where Q(X) is the objective function used to calculate the delay time rate of the affected RC flow; X is a c-dimensional vector representing the particle position, X = [x1, x2, ..., x m ]; c is the number of new TT flows; Indicates the number of affected RC flows; represents the worst end-to-end delay of the ith RC flow, where vl i represents the route of the i-th RC flow, e αβ Represents the connected network node v α 、v β The directed edge of Indicates that any frame of the RC flow is from network node v α Transmit to network node v β The worst time interval; dl i represents the end-to-end delay of the i-th RC flow; D ave represents the average delay time rate of all RC flows, F RC |Indicates the total number of RC flows in the network.

[0027] Furthermore, the penalty function is expressed as:

[0028]

[0029] Where E represents the set of directed communication links between network topology nodes; vl j represents the route of the new TT flow or RC flow in the network described in the jth item; e represents the directed communication link between the network topology nodes; l j p represents the data stream frame length of the new TT stream or RC stream in the network described in item j; j represents the data flow period of the new TT flow or RC flow in the network described in item j; C j Indicates the maximum bandwidth service rate of the directed link; It indicates that the TT flows to be admitted and the RC flows in the network are expected to meet the maximum bandwidth constraint.

[0030] Furthermore, the calculation method of the worst-case end-to-end delay of each RC stream is:

[0031] Calculate the worst transmission interval of each RC flow between every two nodes of the network link based on the preferred routing solution;

[0032] The worst-case delay of the corresponding RC stream is calculated based on all the worst transmission intervals of each RC stream.

[0033] Further, the calculation of the worst transmission interval of each RC flow between every two nodes in the network link based on the routing scheme includes:

[0034]

[0035] in, Indicates that any frame of the RC flow is from network node v α Transmit to network node v β The worst time interval; and Respectively represent any frame of the RC flow transmitted to the network node v β and v α moment; represents the technical delay introduced by each data frame; represents the delay effect of the TT flow on the RC flow, where F TT represents a set of other TT streams other than the new TT stream, f k Indicates other TT flows other than the new TT flow. Indicates that any frame of the RC flow is from the network node v α Transmit to network node v β The time offset between k represents the data stream frame time length of the new TT stream described in the kth item, p k represents the new TT data flow period described in Article k; Indicates the delay impact of other RC flows on this RC flow, where F RC represents the RC flow set, f l represents the new TT flow, F TTRec represents the new TT flow set, l l Indicates the time length of the lth other RC stream frame, p k It indicates the lth other RC flow cycle; It indicates the additional delay of TT flow to the RC due to the preemptive integration strategy; l i Indicates the frame time length of the RC stream.

[0036] Further, the acquiring of multiple optional routing solutions based on the received admission request of the new TT flow set includes:

[0037] Based on the admission request of the received new TT flow set and the network topology, multiple optional routing solutions are calculated using the Dijkstra algorithm.

[0038] Further, selecting a second routing scheme from the optional routing schemes in which the number of RC flows whose end-to-end delay exceeds the corresponding worst delay is as small as possible includes:

[0039] The second routing solution is iteratively solved based on the preferred routing solutions. In each round of iteration:

[0040] Selecting a preferred routing scheme with the least number of RC flows whose end-to-end delay exceeds the corresponding worst delay among the preferred routing schemes;

[0041] Solve the second scheduling solution based on the optimal routing solution and scheduling constraints:

[0042] If successful, the preferred routing solution becomes the second routing solution, and the iteration ends;

[0043] If it fails, continue to the next iteration.

[0044] The present invention can achieve at least one of the following beneficial effects:

[0045] By considering the delay time rate and maximum bandwidth constraint of the affected RC flow, the admission rules for the new TT flow are determined, including the routing plan and scheduling plan. The impact of the new TT flow to be admitted in the TSN network on the RC flow is minimized by dynamic planning, which greatly reduces the probability of the new TT flow affecting the end-to-end delay of the RC traffic, and avoids the subsequent secondary reconstruction of the RC flow from the source as much as possible.

[0046] Through particle swarm optimization algorithm modeling, according to the combination of different routing schemes for the new TT flows to be admitted, considering whether the worst-case end-to-end delay of each RC flow does not exceed the corresponding end-to-end delay limit, the optimal solution of the routing and scheduling schemes for the new TT flows that meet the requirements can be quickly solved.

[0047] Other features and advantages of the present invention will be described in the following description, and some advantages may become apparent from the description, or may be understood by practicing the present invention. The purpose and other advantages of the present invention may be realized and obtained through the contents particularly pointed out in the description, claims and drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] The accompanying drawings are only used for the purpose of illustrating specific embodiments and are not to be considered as limiting the present invention. In the entire drawings, the same reference symbols represent the same components;

[0049] Figure 1 The figure is a flow chart of the method of the present invention. DETAILED DESCRIPTION

[0050] The preferred embodiments of the present invention are described in detail below in conjunction with the accompanying drawings, wherein the accompanying drawings constitute a part of this application and are used together with the embodiments of the present invention to illustrate the principles of the present invention, but are not used to limit the scope of the present invention.

[0051] A specific embodiment of the present invention discloses a bus network time-triggered traffic admission planning method taking into account rate-limited traffic, which specifically includes steps S01 to S03:

[0052] Step S01: based on the received admission request of the new TT flow set, obtain multiple optional routing solutions one by one;

[0053] Step S02: when there is a first routing scheme in the optional routing schemes that satisfies that the end-to-end worst-case delay of each RC flow does not exceed the corresponding end-to-end delay limit, and there is a first scheduling scheme based on the first routing scheme that satisfies the scheduling constraint, execute the first scheduling scheme; otherwise, proceed to step S03;

[0054] Step S03: Select a second routing scheme from the optional routing schemes in which the number of RC flows whose end-to-end worst-case delay exceeds the corresponding end-to-end delay limit is as small as possible, and if a second scheduling scheme that satisfies the scheduling constraint exists based on the second routing scheme, execute the second scheduling scheme.

[0055] The present embodiment discloses a bus network time-triggered traffic admission planning method that takes into account rate-limited traffic. The admission rules for new TT flows are determined by considering the impact of the new TT flows to be admitted on the end-to-end delay of the RC flows in the network, including routing schemes and scheduling schemes. The impact of the dynamic planning of the new TT flows to be admitted in the TSN network on the RC flows is minimized, and the probability of the new TT flows affecting the end-to-end delay of the RC flows is reduced as much as possible.

[0056] It should be noted that the bus network time-triggered traffic admission planning method that takes into account rate-limited traffic described in the present invention adopts a TAS scheduling model based on IEEE 802.1Qbv, and its traffic includes time-triggered (TT) traffic, rate-limited (RC) traffic (such as audio and video AVB traffic) and traditional Ethernet best effort (BE) traffic. Furthermore, in the present invention, only TT traffic and RC traffic are considered. Without loss of generality, the RC traffic in the present invention is only sent periodically in the source system, and no additional scheduling algorithm is executed in the port of the switch passing through, and only the basic first-in, first-out principle (FIFO) is followed. This assumption is reasonable and is backward compatible with more complex scheduling algorithms, so it has the widest range of applications.

[0057] Specifically, the TSN network topology is represented as a directed graph G(V,E), where V represents the set of network nodes of the terminal system (ES) and the switch (SW), and E represents the set of directed communication links between the network nodes. Let v1 and v2 represent two adjacent nodes, and satisfy v1,v2∈V, then [v1,v2] and [v2,v1] can be used to represent bidirectional data flow links. Edge e st It can be described as e st =[v s ,v t ]∈E,v s ,v t ∈V. The multi-priority traffic in time-sensitive networking (TSN) mainly includes TT, rate-limited (RC) and best-effort (BE) traffic. TT traffic is forwarded according to the offline scheduled schedule and has the highest priority. RC traffic is only sent during the execution period of the source system and is subject to the rate limit of the bandwidth allocation gap (BAG). Its priority is lower than TT traffic. BE traffic is traditional Ethernet traffic without any QoS guarantee and has the lowest priority, which is not discussed in this invention.

[0058] Furthermore, all communication tasks in the network are transmitted through data streams, including data frame instances. Let F = F TT ∪F RC Represents the set of all data flows. Any TT flow f m ∈F TT and any RC flow f n ∈F RC They can be represented by the following seven-tuple and six-tuple respectively:

[0059] f m ={s m ,d m ,l m ,p m ,dl m ,vl m ,O m} and f n ={s n ,d n ,l n ,p n ,dl n ,vl n};

[0060] where s m d m , l m 、p m 、dl m and vl mRespectively represent the source ES, destination ES, frame length in the time dimension (frame length divided by link bandwidth), traffic period, end-to-end (ETE) delay limit, and routing information. m Exclusive to TT traffic, it represents the offset set between the first data frame instance and the start time of the cycle of the TT flow on each directed link. The offset set of all TT flows together constitutes the time schedule of the network. In simple terms, the admission control and dynamic planning of TT flows means reallocating the routing information and corresponding offset time of TT flows.

[0061] Specifically, in one embodiment of the present invention, in step S01, when an admission request of a new TT flow set is received, multiple optional routing solutions are obtained according to the current TSN network topology.

[0062] Specifically, based on the admission request of the received new TT flow set and the network topology, the Dijkstra algorithm is used to calculate k optional routing solutions (such as Figure 1 As shown in the reference numeral 1), since the use of Dijkstra algorithm to solve the optional routing solution is a prior art, the present invention will not elaborate on it.

[0063] Further, step S02 includes steps S021 to S022:

[0064] S021, determining whether there is a first routing solution among the optional routing solutions that satisfies that the end-to-end worst-case delay of each RC flow does not exceed the corresponding end-to-end delay limit, if yes, proceeding to step S022, if not, proceeding to step S03;

[0065] S022, solving a first scheduling solution that satisfies the scheduling constraints based on the first routing solution (such as Figure 1 If the solution is successful (as shown in the label 4), Figure 1 5), then execute the first scheduling scheme (such as Figure 1 If the solution fails, go to step S03.

[0066] Furthermore, step S021 includes:

[0067] With the goal of optimizing the delay time rate of each RC flow affected by the new TT flow, an optimal routing solution (such as Figure 1 (shown in the middle label 2);

[0068] In each round of iteration, based on the optimal routing solution obtained in this round of iteration, it is determined whether the worst-case end-to-end delay of each RC flow does not exceed the corresponding end-to-end delay limit (such as Figure 1 (as shown in the bid number 3):

[0069] If yes, the optimal routing solution obtained in this round of iteration is the first routing solution;

[0070] If not, continue to the next iteration (such as Figure 1 2 and 3), until the maximum number of iterations is reached (as shown in Figure 1 (shown in the middle label 8).

[0071] Furthermore, in step S021, a particle swarm algorithm is used to iteratively solve an optimal routing solution based on the optional routing solution. The process of using the particle swarm algorithm to solve the optimal routing solution includes:

[0072] step-a, numbering the k optional routing solutions;

[0073] step-b, taking the combination of the routing schemes of the new TT flows as particles, describing the particle positions with the combination of the corresponding numbers of the routing schemes of the new TT flows in each particle, and randomly initializing the particle swarm based on the multiple optional routing schemes;

[0074] Step-c, construct a fitness function based on the delay time rate and maximum bandwidth constraint of the affected RC flow;

[0075] step-d: In each round of iteration, the optimal solution for the particle swarm is sought based on the fitness function, and the optimal routing solution for this round of iteration is obtained and recorded.

[0076] Furthermore, in step-a, the selectable routing scheme number ranges from 1 to k.

[0077] Furthermore, in step-b, the position of each particle is represented by an M-dimensional vector X = [x1, x2, ..., x M ], where x m (m∈[1,M]) represents the new TT flow f m The routing scheme number is an integer ranging from 1 to k. The speed of the particle is expressed as V = [v1, v2, ..., v M ], where v m (m∈[1,M]) represents the distance the particle moves in the mth dimension, ranging from -(k-1) to (k-1).

[0078] Furthermore, in step-c, the fitness function is constructed based on the delay time rate and the maximum bandwidth constraint of the affected RC flow, including:

[0079] An objective function is constructed based on the delay time rate of the affected RC flows;

[0080] Construct a penalty function based on the maximum bandwidth constraint;

[0081] The fitness function is constructed based on the objective function, the penalty function and the weight coefficient of the penalty function.

[0082] Furthermore, the objective function is expressed as:

[0083]

[0084] Where Q(X) is the objective function used to calculate the delay time rate of the affected RC flow; X is a c-dimensional vector representing the particle position, X = [x1, x2, ..., x m ]; c is the number of new TT flows; Indicates the number of affected RC flows; represents the worst end-to-end delay of the ith RC flow, where vl i represents the route e of the i-th RC flow αβ Represents the connected network node v α 、v β The directed edge of Indicates that any frame of the RC flow is from network node v α Transmit to network node v β The worst time interval; dl i represents the end-to-end delay of the i-th RC flow; D ave represents the average delay time rate of all RC flows, F RC |Indicates the total number of RC flows in the network.

[0085] Furthermore, the penalty function is expressed as:

[0086]

[0087] Where E represents the set of directed communication links between network topology nodes; vl j represents the route of the new TT flow or RC flow in the network described in the jth item; e represents the directed communication link between the network topology nodes; l j p represents the data stream frame length of the new TT stream or RC stream in the network described in item j; j represents the data flow period of the new TT flow or RC flow in the network described in item j; C j Indicates the maximum bandwidth service rate of the directed link; It indicates that the TT flows to be admitted and the RC flows in the network are expected to meet the maximum bandwidth constraint.

[0088] Furthermore, a fitness function is constructed based on the objective function, the penalty function and the weight coefficient of the penalty function to evaluate the performance of the routing scheme represented by the particle position. The lower the fitness function value, the better the performance of the selected route.

[0089] The fitness function is expressed as: F(X)=Q(X)+γE(X); wherein γ is a penalty weight, which varies according to different topologies and traffic flows, and preferably, is 300.

[0090] Furthermore, in step-d, the optimal solution of the particle swarm is sought based on the fitness function. Specifically, in the process of searching for the optimal solution, the position and velocity of the particle are updated according to two values: the current optimal solution pBest of a single particle and the optimal solution gBest of all particles in the swarm; the search process is expressed as:

[0091] V t+1 =ωV t +c1r1(pBest-X t )+c2r2(gBest-X t );

[0092] X t+1 =X t +V t ;

[0093] Among them, ω is the inertia weight, which is used to adjust the degree to which the particle maintains its original speed. The value range is between 0 and 1, and the preferred setting is 0.9; c1 and c2 represent self-learning and social learning factors, respectively, which are used to balance local and global search capabilities. c1 controls the speed and direction of self-update of particles in the search space. The larger the value, the more inclined to search along the direction of its own historical optimal solution, and the smaller the value, the more inclined to random exploration; c2 controls the speed at which particles update in space according to the optimal solution of neighboring particles. The larger the value, the more inclined to search under the influence of neighboring particles, and the smaller the value, the more inclined to independent search; c 1 The sum of c1 and c2 is not greater than 4. Preferably, c1 and c2 are set to 2 and 0.9 respectively; r1 and r2 are random numbers between 0 and 1; t represents the number of searches in the solution process.

[0094] Furthermore, in step S021, in each round of iteration, it is determined based on the optimal routing solution obtained in the round of iteration whether the end-to-end worst-case delay of each RC flow does not exceed the corresponding end-to-end delay limit.

[0095] The calculation method of the worst-case end-to-end delay of each RC stream is:

[0096] Calculate the worst transmission interval of each RC flow between every two nodes of the network link based on the preferred routing solution;

[0097] The worst-case delay of the corresponding RC stream is calculated based on all the worst transmission intervals of each RC stream.

[0098] Further, the calculation of the worst transmission interval of each RC flow between every two nodes in the network link based on the preferred routing solution includes:

[0099]

[0100] in, Indicates that any frame of the RC flow is from network node v α Transmit to network node v β The worst time interval; and Respectively represent any frame of the RC flow transmitted to the network node v β and v α moment; represents the technical delay introduced by each data frame; represents the delay effect of the TT flow on the RC flow, where F TT represents a set of other TT streams other than the new TT stream, f k Indicates other TT streams other than the new TT stream, Any frame of the RC flow is sent from the network node v α Transmit to network node v β The time offset between k represents the data stream frame time length of the new TT stream described in the kth item, p k represents the new TT data flow period described in Article k; Indicates the delay impact of other RC flows on this RC flow, where F RC represents the RC flow set, f l represents the new TT flow, F TTRec represents the new TT flow set, l l Indicates the time length of the lth other RC stream frame, p k It indicates the lth other RC flow cycle; It indicates the additional delay of TT flow to the RC due to the preemptive integration strategy; l i Indicates the frame time length of the RC stream.

[0101] The following describes the principle of calculating the worst transmission interval of each RC flow between every two nodes in the network link:

[0102] In the method of the present invention, it is reasonably assumed that all RC flows have passed the worst-case end-to-end delay verification in the design phase, i.e. (wcd nrepresents the worst-case end-to-end delay of the nth RC flow). When judging whether the worst-case end-to-end delay of each RC flow does not exceed the corresponding end-to-end delay limit, it is only necessary to judge the RC flows that have an intersection with the routes of the new TT flow. This is because in delay analysis, the end-to-end delay of non-intersecting RC flows may change, but their worst-case end-to-end delay wcd remains unchanged (regardless of delay calculation methods such as network calculation or trajectory method).

[0103] Furthermore, the end-to-end delay of the RC flow is mainly affected by four aspects in addition to the transmission delay: ① related technical delay, such as the delay caused by the data validity and integrity check between nodes, the delay caused by the bandwidth allocation interval error containment delay, etc.; ② the delay caused by the competition for bandwidth resources between RC flows; ③ the delay caused by high-priority TT traffic to the RC flow; ④ the delay caused by the integration of TT and RC flows, which is mainly divided into three types according to different integration strategies: shuffling, preemption and timely blocking. The most common preemption strategy (this strategy is the most general, representative and practical) is adopted in the method of the present invention.

[0104] Furthermore, the end-to-end delay of the RC flow mainly consists of two parts: ① the queuing delay of the source ES; ② the delay of each intermediate node (link).

[0105] First, consider the queuing delay at the source ES. Specifically, the set of RC flows that have an intersection with the routing of the new TT flow is identified as F RC ′. Since the initial phase (offset) of the route corresponding to the RC flow can be allocated by the traditional offset allocation algorithm, vl i The initial offset is recorded as Furthermore, the transmission time when it leaves the end node and enters the network (denoted as ) can be calculated as follows:

[0106]

[0107] Where FIFO(j,i) represents TT flow f j The data frame is earlier than the RC stream f i The data frame enters the buffer; when FIFO(j,i) is an empty set, Therefore, the RC flow f entering the network i The offset can be represented by several possible discrete values.

[0108] Secondly, the delay calculation of each intermediate node (link) is considered. Specifically, the present invention adopts an iterative strategy of overall timing analysis and uses the concept of busy period to analyze the delay of the affected RC flow in the network. For any data frame of the RC flow, its busy period on a certain link can be defined as a time interval, that is, from the time when the frame arrives at the node v α To transmit the frame to the next node v β The time interval is expressed as Define the frame arrival node v α To transmit the frame to the next node v β Time interval The end-to-end delay of the RC flow is expressed as To ensure that the worst-case end-to-end delay of each RC flow does not exceed the corresponding end-to-end delay limit dl i , the worst time interval Described as (for any known ):

[0109]

[0110] in, represents the technical delay introduced by each data frame; Indicates the delay impact of the TT flow on the RC flow; Indicates the delay impact of other RC flows on this RC flow; It indicates the additional delay of TT flow to this RC due to the preemptive integration strategy.

[0111] Since the above formula contains both sides is a recursive function, so we can use the recursive tree method to solve it. Indicates from At the beginning of time, link e αβ Medium RC flow i The worst-case end-to-end delay, which guarantees the transmission of all TT flows and other backlogged RC flows that arrive earlier within the time period. Can be used as vl i The start time of the next link in the process is used to solve the busy period until the until.

[0112] The above is the principle of calculating the worst transmission interval of each RC flow between every two nodes in the network link.

[0113] Further, in step S021, it is determined whether the worst-case end-to-end delay of each RC stream does not exceed the corresponding end-to-end delay limit, which can be expressed as:

[0114]

[0115] Furthermore, in step S022, an integer linear programming solver is used to solve the first scheduling solution satisfying the scheduling constraint based on the first routing solution, and the solver may be gurobi, cplex, etc. Since the solver and the solving method are both prior art, they will not be discussed here.

[0116] Further, if the first scheduling solution is successfully solved based on the first routing solution, the first scheduling solution (such as Figure 1 As shown in the middle label 6) as a new configuration scheme A (such as Figure 1 7) is distributed to the corresponding nodes in the network for configuration switching to execute the first scheduling scheme. If it fails, it is determined whether the maximum number of iterations of the particle swarm algorithm has been reached. If the maximum number of iterations has not been reached, step S021 is re-entered to solve the first routing scheme for the next iteration; if the maximum number of iterations has been reached, it means that the solution of the first scheduling scheme has failed, and step S03 is entered. It should be noted that, in general, solving the first scheduling scheme based on the first routing scheme is related to the network load, and the probability of failure in practical applications is extremely small.

[0117] Further, in step S03, based on all the optimal routing solutions obtained by the particle swarm algorithm in step S02, the second routing solution is iteratively solved. Specifically, in each round of iteration:

[0118] Select the preferred routing scheme with the least number of RC flows whose end-to-end delay exceeds the corresponding worst delay among the preferred routing schemes (such as Figure 1 (shown in the middle label 9);

[0119] Based on the preferred routing solution and the scheduling constraints, a second scheduling solution (such as Figure 1 (as shown in the bid number 10):

[0120] If successful (such as Figure 1 11), the preferred routing solution is the second routing solution, the iteration ends, and the second scheduling solution is dispatched (as shown in FIG. Figure 1 As shown in the middle label 13) as a new configuration scheme B (such as Figure 1 The node (shown in the reference numeral 14) is distributed to the corresponding node in the network to perform configuration switching to execute the second scheduling scheme;

[0121] If it fails, determine whether all the preferred routing solutions have been traversed (such as Figure 1 If not, continue to the next iteration (as shown in the label 12). Figure 1 If yes, it means that the second scheduling solution fails and the new TT flow admission request is not passed (such as Figure 1 (shown in the reference numeral 15).

[0122] It should be noted that, since in each round of iteration, solving the second scheduling solution based on the preferred routing solution and scheduling constraints selected in that round is related to the network load, in practical applications, the probability of failure in solving the second scheduling solution is extremely small.

[0123] The present embodiment discloses a bus network time-triggered traffic admission planning method that takes into account rate-limited traffic. By considering the delay time rate and maximum bandwidth constraint of the affected RC flow, the admission rules for new TT flows including routing schemes and scheduling schemes are determined, and the impact of new TT flows to be admitted in the TSN network on the RC flow is dynamically planned to be minimized, thereby solving the problem of the new TT flow affecting the end-to-end delay of the RC flow.

[0124] Through particle swarm optimization algorithm modeling, according to the combination of different routing schemes for the new TT flows to be admitted, considering whether the worst-case end-to-end delay of each RC flow does not exceed the corresponding end-to-end delay limit, the optimal solution of the routing and scheduling schemes for the new TT flows that meet the requirements can be quickly solved.

[0125] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by any technician familiar with the technical field within the technical scope disclosed by the present invention should be covered within the protection scope of the present invention.

Claims

1. A bus network time-triggered traffic admission planning method taking into account rate-limited traffic, characterized in that: The steps include: Based on the admission request of the received new TT flow set, multiple optional routing solutions are obtained one by one; When there is a first routing scheme in the optional routing schemes that satisfies that the end-to-end worst-case delay of each RC flow does not exceed the corresponding end-to-end delay limit, and there is a first scheduling scheme based on the first routing scheme that satisfies the scheduling constraint, executing the first scheduling scheme; otherwise, A second routing scheme in which the number of RC flows whose end-to-end worst-case delay exceeds the corresponding end-to-end delay limit is as small as possible is selected from the optional routing schemes, and when a second scheduling scheme that satisfies the scheduling constraint exists based on the second routing scheme, the second scheduling scheme is executed.

2. The bus network time-triggered traffic admission planning method according to claim 1 is characterized in that: Determining whether there is a first routing solution among the optional routing solutions that satisfies that the end-to-end worst-case delay of each RC flow does not exceed the corresponding end-to-end delay limit includes: Iteratively solving an optimal routing solution based on the optional routing solution with the goal of optimizing the delay time rate of each RC flow affected by the new TT flow; In each round of iteration, based on the optimal routing solution obtained in this round of iteration, it is determined whether the worst-case end-to-end delay of each RC flow does not exceed the corresponding end-to-end delay limit: If yes, the optimal routing solution obtained in this round of iteration is the first routing solution; If not, continue with the next iteration until the maximum number of iterations is reached.

3. The bus network time-triggered traffic admission planning method according to claim 2 is characterized in that: Using a particle swarm algorithm to optimize the delay time rate of each RC flow affected by the new TT flow, iteratively solving an optimal routing solution based on the optional routing solution, including: Numbering the optional routing solutions; A combination of the routing schemes of the new TT flows is used as a particle, a combination of corresponding numbers of the routing schemes of the new TT flows in each particle is used to describe the particle position, and a particle group is randomly initialized based on the multiple optional routing schemes; A fitness function is constructed based on the delay time rate and maximum bandwidth constraint of the affected RC flows; In each round of iteration, the optimal solution of the particle swarm is sought based on the fitness function to obtain the optimal routing solution for that round of iteration.

4. The bus network time-triggered traffic admission planning method according to claim 3 is characterized in that: The fitness function constructed based on the delay time rate and maximum bandwidth constraint of the affected RC stream includes: An objective function is constructed based on the delay time rate of the affected RC flows; Construct a penalty function based on the maximum bandwidth constraint; The fitness function is constructed based on the objective function, the penalty function and the weight coefficient of the penalty function.

5. The bus network time-triggered traffic admission planning method according to claim 4 is characterized in that: The objective function is expressed as: Where Q(X) is the objective function used to calculate the delay time rate of the affected RC flow; X is a c-dimensional vector representing the particle position, X = [x1, x2, ..., x m ]; c is the number of new TT flows; Indicates the number of affected RC flows; represents the worst end-to-end delay of the ith RC flow, where vl i represents the route of the i-th RC flow, e αβ Represents the connected network node v α 、v β The directed edge of Indicates that any frame of the RC flow is from network node v α Transmit to network node v β The worst time interval; dl i represents the end-to-end delay of the i-th RC flow; D ave represents the average delay time rate of all RC flows, |F RC |Indicates the total number of RC flows in the network.

6. The bus network time-triggered traffic admission planning method according to claim 5, characterized in that: The penalty function is expressed as: Where E represents the set of directed communication links between network topology nodes; vl j represents the route of the new TT flow or RC flow in the network described in the jth item; e represents the directed communication link between the network topology nodes; l j p represents the data stream frame length of the new TT stream or RC stream in the network described in item j; j represents the data flow period of the new TT flow or RC flow in the network described in item j; C j Indicates the maximum bandwidth service rate of the directed link; It indicates that the TT flows to be admitted and the RC flows in the network are expected to meet the maximum bandwidth constraint.

7. The bus network time-triggered traffic admission planning method according to claim 2, characterized in that: The calculation method of the worst-case end-to-end delay of each RC stream is: Calculate the worst transmission interval of each RC flow between every two nodes of the network link based on the preferred routing solution; The worst-case delay of the corresponding RC stream is calculated based on all the worst transmission intervals of each RC stream.

8. The bus network time-triggered traffic admission planning method according to claim 7, characterized in that: The calculation of the worst transmission interval of each RC flow between every two nodes in the network link based on the routing scheme includes: in, Indicates that any frame of the RC flow is from network node v α Transmit to network node v β The worst time interval; and Respectively represent any frame of the RC flow transmitted to the network node v β and v α moment; represents the technical delay introduced by each data frame; represents the delay effect of the TT flow on the RC flow, where F TT represents a set of other TT streams other than the new TT stream, f k Indicates other TT flows other than the new TT flow. Any frame of the RC flow is sent from the network node v α Transmit to network node v β The time offset between k represents the data stream frame time length of the new TT stream described in the kth item, p k represents the new TT data flow period described in Article k; Indicates the delay impact of other RC flows on this RC flow, where F RC represents the RC flow set, f l represents the new TT flow, F TTRec represents the new TT flow set, l l Indicates the time length of the lth other RC stream frame, p k It indicates the lth other RC flow cycle; It indicates the additional delay of TT flow to the RC due to the preemptive integration strategy; l i Indicates the frame time length of the RC stream.

9. The bus network time-triggered traffic admission planning method according to any one of claims 1 to 8, characterized in that: The acquiring of multiple optional routing solutions based on the received admission request of the new TT flow set includes: Based on the admission request of the received new TT flow set and the network topology, multiple optional routing solutions are calculated using the Dijkstra algorithm.

10. The bus network time-triggered traffic admission planning method according to claim 2, characterized in that: Selecting a second routing solution from the optional routing solutions, in which the number of RC flows with the end-to-end delay exceeding the corresponding worst delay is as small as possible, comprises: The second routing solution is iteratively solved based on the preferred routing solutions. In each round of iteration: Selecting a preferred routing scheme with the least number of RC flows whose end-to-end delay exceeds the corresponding worst delay among the preferred routing schemes; Solve the second scheduling solution based on the optimal routing solution and scheduling constraints: If successful, the preferred routing solution becomes the second routing solution, and the iteration ends; If it fails, continue to the next iteration.

Images