CN117278487A - Flow scheduling method of avionic system based on Qbv protocol of time sensitive network - Google Patents

Flow scheduling method of avionic system based on Qbv protocol of time sensitive network Download PDF

Info

Publication number
CN117278487A
CN117278487A CN202311375303.7A CN202311375303A CN117278487A CN 117278487 A CN117278487 A CN 117278487A CN 202311375303 A CN202311375303 A CN 202311375303A CN 117278487 A CN117278487 A CN 117278487A
Authority
CN
China
Prior art keywords
flow
scheduling
link
current
traffic
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202311375303.7A
Other languages
Chinese (zh)
Inventor
周璇
何锋
景世龙
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Beihang University
Original Assignee
Beihang University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Beihang University filed Critical Beihang University
Priority to CN202311375303.7A priority Critical patent/CN117278487A/en
Publication of CN117278487A publication Critical patent/CN117278487A/en
Pending legal-status Critical Current

Links

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L47/00—Traffic control in data switching networks
    • H04L47/50—Queue scheduling
    • H04L47/62—Queue scheduling characterised by scheduling criteria
    • H04L47/6245—Modifications to standard FIFO or LIFO
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L45/00—Routing or path finding of packets in data switching networks
    • H04L45/38—Flow based routing
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L45/00—Routing or path finding of packets in data switching networks
    • H04L45/54—Organization of routing tables
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L47/00—Traffic control in data switching networks
    • H04L47/50—Queue scheduling
    • H04L47/62—Queue scheduling characterised by scheduling criteria
    • H04L47/625—Queue scheduling characterised by scheduling criteria for service slots or service orders
    • H04L47/6275—Queue scheduling characterised by scheduling criteria for service slots or service orders based on priority

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Data Exchanges In Wide-Area Networks (AREA)

Abstract

The invention discloses a flow scheduling method of an avionic system based on a Qbv (quantum-based virtual bus) protocol of a time-sensitive network, which comprises the steps of initializing flow scheduling, acquiring flow and a transmission path thereof, calculating an available time slot range boundary, eliminating scheduling invalid time, calculating flow scheduling time, updating a flow scheduling queuing sequence, generating a scheduling Gantt chart and synthesizing a gating list. The method of the invention directly calculates the bias for the dispatching flow in the TSN network by utilizing the modes of dynamic sequencing and boundary conversion, thereby not only overcoming the dependence of the traditional constraint solving method on an external solver, but also avoiding the randomness of iterative optimization in the intelligent optimizing method, and further rapidly generating a deterministic flow dispatching table serving in an avionics system.

Description

Flow scheduling method of avionic system based on Qbv protocol of time sensitive network
Technical Field
The invention relates to an avionic system applying a time-sensitive network Qbv protocol, in particular to a flow scheduling method of the avionic system based on the time-sensitive network Qbv protocol, which is quick flow scheduling based on flow dynamic sequencing and boundary equivalence conversion.
Background
Avionics systems are evolving towards deep integration, and higher demands are being put on capacity, real-time performance, reliability and the like of data transmission. Time sensitive networks (Time-Sensitive Networking, TSNs) have attracted considerable attention in the aerospace industry as current "on-Time, accurate" end-to-end transmission candidate protocols. In the TSN protocol cluster, the IEEE 802.1Qbv protocol proposes a Time-Aware shaping (TAS) flow control mechanism based on a gating list (Gate Control List, GCL), which can ensure the transmission of high-priority schedulable traffic (Scheduled Traffic, ST) by controlling the on and off times of Time windows of different priority traffic queues. Designing a viable traffic scheduling scheme and GCL has become a prerequisite for applying the Qbv protocol, but the switched interconnect and First-In-First-Out (FIFO) principle of TSN networks increases the complexity of traffic scheduling constraints, which brings great challenges to scheduling design.
The existing related scheduling method mostly adopts a constraint solving method and an intelligent optimizing method. The constraint solving method is mainly used for carrying out planning solving by combining a generated network topological structure and flow configuration and converting constraint conditions into a mathematical logic expression of a satisfiability modulus theory (Satisfiability Modulo Theories, SMT) or an integer linear programming (Integer Linear Programming, ILP), and the scheduling design of the constraint solving method is usually finished by carrying out intensive operation for a long time by depending on a corresponding solver. The intelligent optimization method is used for realizing dispatching optimization design by setting an optimization target and a search strategy by referring to the thought of heuristic learning or machine learning, but the dispatching scheme has randomness due to the random selection of iterative search. As can be seen, current avionics systems still lack a method to quickly generate deterministic schedules for TSN networks.
Disclosure of Invention
Aiming at the flow scheduling under the time-sensitive network Qbv protocol applied in the avionic system, the existing constraint solving method or intelligent optimizing method still has the defects of dependence on an external solver, longer solving time or random scheduling scheme and the like. The flow scheduling method of the invention converts complex constraint conditions among flows into a usable time slot range of the flows, dynamically updates the flow scheduling queuing sequence in real time, thereby rapidly generating a deterministic flow scheduling table, and utilizing the scheduling table to perform network flow scheduling under a time-sensitive network Qbv protocol. When the flow scheduling method is applied to an avionic system, the efficiency of flow scheduling design can be improved, and the certainty of a flow scheduling table can be ensured.
When the network topology structure generated by each aviation device in the avionic system runs the rapid flow scheduling based on the time-sensitive network Qbv protocol, a deterministic flow scheduling table can be rapidly generated directly in a flow dynamic sequencing and boundary equivalence conversion mode without depending on an external solver. In the present invention, the length and period of the same traffic under the time sensitive network Qbv protocol are fixed. Flow m i Length of (2) is recorded asFlow m i The period of (2) is recorded as +.>In the case of a fast calculation of the transmission time +.>And when generating a TSN network available gating list GCL, executing a traffic scheduling method of an avionic system based on a time-sensitive network Qbv protocol as shown in figure 1, wherein the traffic scheduling method comprises the following steps:
step one, initializing flow scheduling;
step 1-1, initializing the scheduling states of all scheduling traffic;
traffic set to be schedulableThe scheduling status of each flow in (1)>Initialized to the unscheduled state undone.
In the present invention,indicating the scheduling status. Said->Including unscheduled tunes and scheduled tunes.
The undne indicates that all traffic in the TSN network is not scheduled in the initial state, which is simply referred to as unscheduled.
done indicates that traffic has been scheduled during the scheduling process.
Step 1-2, initializing the idle time slot range of all links;
in the present invention, the static routing table TAB configured in the TSN network is used Routing And link l (node β ,node χ ) Is of the super period of (a)The free time slot set +.>Initialized to->
Static routing table
Link Traffic transmitted over links
In the invention, the static routing table is a database in the form of a two-column multi-row table and is used for recording the corresponding relation between links and traffic of the avionic system under the time-sensitive network Qbv protocol.
Step 1-3, queuing of flow scheduling;
traffic set to be schedulableAnd sequencing according to the flow period from small to large to obtain a first sequencing set, which is marked as MIA.
And sequencing the flows with the same period in the first sequencing set MIA according to the flow length from small to large to obtain a second sequencing set, and marking the second sequencing set as QM.
Step 1-4, initializing the current dispatching flow;
the current dispatching flow is recorded as cur QM . At initialization, cur QM Is empty.
Initializing, selecting the 1st queuing flow in the second ordered set QM, and marking the 1st queuing flow as m i And subjecting said m to i As the current dispatch traffic, assign to cur QM Then there is cur QM ＝<m i >。
Step two, acquiring the current scheduling flow and a transmission path thereof;
step 2-1, obtaining current dispatching flow;
cur according to the current scheduling flow QM Obtaining traffic to be scheduled, i.e. m i 。
Step 2-2, obtaining a transmission path of the current scheduling flow;
according to static routing table TAB configured in TSN network Routing Obtaining m i Is set of transmission paths of (a)
Step 2-3, initializing the transmission link number of the current dispatching flow;
the current link is denoted as cur QP . At initialization, cur QP Is empty.
Initializing and selectingThe 1st path link->And add said->Current link, cur, as current scheduled traffic QP Assignment of +.>
Step 2-4, obtaining a transmission link of the current scheduling flow;
according to the current link cur QP Obtaining traffic-links to be scheduled, i.e.
Step three, calculating the upper bound and the lower bound of the available time slot range of the current dispatching flow on the current transmission link;
step 3-1, determining a calculation method of upper and lower boundaries of a time slot range;
traffic m to be scheduled i On traffic-link to be scheduledThe available time slot range is recorded asAnd-> Is thatIs defined as the lower boundary value of (a). />Is->Is not included in the upper boundary value of (a).
JudgingWhether or not it is->The first link of (a); if->Is->The first link in (i.e.)>If the source node is the end system node, executing the step 3-2; if->Not->Step 3-3 is performed. In the present invention, step 3-3 includes 4 sub-steps.
Step 3-2, calculating an upper boundary value and a lower boundary value of an available time slot range of the current scheduling flow on the first link;
will beAssignment as free time slot set->Is defined by the left boundary of (c). Calculated according to formula (1)Executing step four. In the present invention, step four includes 4 sub-steps.
Is the traffic m to be scheduled i Is a periodic one. />Is the traffic m to be scheduled i Is a length of (c).
Step 3-3, calculating an upper boundary value and a lower boundary value of an available time slot range of the current scheduling flow on a non-first link;
step 3-3A, findEquivalent Link set->
Step 3-3B according toStatistics m i At->Up-transmission equivalent traffic set
Step 3-3C, inFinding m in i To->The previous linkSaid->Destination node and->Is the same as the source node of (a). Calculate +.>Calculate +.>
Step 3-3D, initializing equivalent flow;
in the present invention, an equivalent flow set is utilizedEquivalent flow in (a) and iteratively updating +.>And->Step four is then performed.
Represents traffic m to be scheduled i Is the number of instances of (a).
Representing the number of instances of equivalent traffic.
Delta is clock synchronization accuracy.
Step four, the invalid moment of the current dispatching flow on the current link is eliminated;
step 4-1, acquiring a subsequent link list of a current link aiming at the current scheduling flow;
at m i Is set of transmission paths of (a)Finding m in i In link->Subsequent Link set on->
Step 4-2, obtaining an interference flow list when the current scheduling flow is transmitted on the current link;
according toStructure m i At->Interference traffic set on->
Step 4-3, determining an invalid time list of the current scheduling flow on the current link;
step 4-3A, judging the interference flow setIf not, executing the step five, and if not, executing the step 4-3B.
Step 4-3B, initializing an invalid time setIs empty;
step 4-3C, utilizeIteratively updating the invalid scheduling time list according to the formula (5);
representing the number of instances of interfering traffic.
Step 4-4, eliminating an invalid time list of the current scheduling flow on the current link;
updating m i At the position ofAvailable time slots ∈>And exclude invalid time sets therefrom
Step five, calculating the scheduling moment of the current scheduling flow on the current transmission link;
step 5-1, judging whether the current dispatching flow can be dispatched on the current transmission link;
according toPreliminary determination of m i At->Whether or not the upper part is schedulable; if it isThen m is i At->The step six is executed after the scheduling is not possible; if->Then m is i At->And (5) performing step 5-2.
Step 5-2, iteratively determining the scheduling time of the current scheduling flow on the current transmission link;
step 5-2A, initializing the iterative determination identifier r, and assigning r asI.e.
Step 5-2B, calculating the estimated occupied time slot of the scheduling flow under the current iteration according to the formula (7)
Step 5-2C, judgingWhether or not it belongs entirely to the free time slot set of the current link +.>If the current scheduling traffic belongs to the current transmission link, the scheduling bias of the current scheduling traffic on the current transmission link can be assigned, namelyLet m i Scheduling state of->Adjust to done, update->And executing the step 5-2D; otherwise, executing the step 5-2G.
Step 5-2D, judging whether the current scheduling flow has completed scheduling calculation on all transmission links, namelyIf the destination node is an end system, executing the step 5-2E, and if not, executing the step 5-2F.
Step 5-2E, updating cur QM Judging whether the scheduling calculation of all the flows is finished, if yes, executing a step seven, otherwise, executing a step 2-1.
Step 5-2F, updating cur QP And executing the step 2-4.
Step 5-2G, judgingIf yes, executing the step six, and if not, executing the step 5-2H.
Step 5-2H, updating the iteration determination identifier r to beThe next available slot in the frame, and returns to step 5-2B.
Step six, dynamically updating a flow scheduling queuing sequence;
step 6-1, determining a blocking flow which causes the current scheduling flow to be non-schedulable on the current link;
calculating an equivalent flow set according to equation (8)Any equivalent flow m y Is the blocking distance of (2)
If m is y Is the smallest, then m y The identity of (a) is changed to m i Is blocking flow.
Step 6-2, updating the idle transmission time slot of the related link;
finding blocked traffic m in queuing sequence QM y And subsequent traffic, compensating for the time slots occupied by these traffic on the corresponding transmission link, updating the free time slots of the associated link
Step 6-3, updating a queuing sequence of flow scheduling;
the traffic m to be scheduled in QM i Inserting blocking flow m y Before, a new queuing sequence QM is formed New type 。
Step 6-4, updating the sequencing number of the current scheduling flow;
updating cur QM Judging whether the step is finished, if so, executing the step 7, and if not, executing the step 2-1.
Step seven, completing the scheduling design of all the flows;
step 7-1, generating a network scheduling Gantt chart;
and converting the scheduling time of all the traffic on the corresponding link into the scheduling time of the output port of the network node under the Qbv protocol of the time-sensitive network, and summarizing the scheduling conditions of all the output ports to form a full-network scheduling Gantt chart.
Step 7-2, synthesizing a gating list of the TSN output port;
and according to the actual occupation condition of each link in the scheduling Gantt chart, combining the time slot length of the gating list GCL to generate an output end gating list GCL of the node output port corresponding to each link.
The flow scheduling method of the avionic system based on the time-sensitive network Qbv protocol has the advantages that: (1) The flow scheduling method can complete flow scheduling planning by only using network topology and flow configuration information generated by each aviation device, can overcome the inherent dependence of the traditional constraint solving method on external solvers such as Gurobi, CPLEX and the like, and can solve the difficult problem of parameter configuration selection in a related optimization algorithm by an intelligent optimization method, thereby realizing scheduling solving with flexibility and certainty. (2) The flow scheduling method can convert complex constraint conditions among flows into the available time slot range of flow scheduling bias, can avoid the problem of long time consumption when directly solving a large number of complex constraint conditions, and can improve the efficiency of flow scheduling solving. (3) The flow scheduling method can dynamically and adaptively update the flow scheduling sequence, can avoid the problem that partial flows cannot be scheduled due to fixed scheduling sequence, and can improve the schedulable flow scale.
Drawings
Fig. 1 is a flow chart of the flow scheduling of the avionics system of the present invention based on the time sensitive network Qbv protocol.
Fig. 2 is a schematic diagram of a time-sensitive network topology conforming to the Qbv protocol.
Fig. 3 is a graph of a scheduled gater obtained by the traffic scheduling method of the present invention.
Fig. 4 is an end-to-end delay result of a scheduling plan obtained by the traffic scheduling method of the present invention.
Detailed Description
The invention will be described in further detail with reference to the drawings and examples.
In the present invention, ST traffic in a TSN network refers to a plurality of schedulable traffic (i.e., ST traffic) in the TSN network expressed in a collective form, denoted as schedulable traffic set MI, andi represents the identification number of the schedulable traffic, I e {1,2, …, I }, I represents the total number of schedulable traffic. m is m 1 Representing the 1st schedulable traffic in a TSN network. m is m 2 Representing the 2 nd schedulable traffic in the TSN network. m is m i Representing the ith schedulable traffic in the TSN network. m is m i-1 Representing the i-1 th schedulable traffic in the TSN network. m is m i+1 Representing the i+1th schedulable traffic in the TSN network. m is m x Representing the x-th schedulable traffic in the TSN network. m is m y Representing the y-th schedulable traffic in the TSN network. m is m y-1 Representing the y-1 th schedulable traffic in the TSN network. m is m y+1 Representing the y+1th schedulable traffic in the TSN network. m is m I Representing the last schedulable traffic in the TSN network. For convenience of explanation, m i Also called any one of the schedulable flows, m y Also referred to as any other schedulable traffic, and schedulable traffic m i And schedulable traffic m y Not the same flow. When m is i Referred to as current schedulable traffic, m i-1 Called m i Is the previous schedulable traffic, m i+1 Called m i The latter one of the schedulable traffic.
In a TSN network, a logical link and a transmission path refer to unidirectional transmission of traffic between two nodes, referred to as a logical link. The transmission of the same traffic over multiple logical links is referred to as a transmission path.
To describe the flow m in detail i A multi-source transmission path from a source node to a plurality of destination nodes is described as follows, in a first aspect, traffic m i Slave node β To node 2 The logical link is marked asFlow m i Slave node 2 To node χ The logical link is marked as +.>Flow m i Slave node χ To node 4 The logical link is marked as +.>Flow m i Slave node 4 To node 3 The logical link is marked as +.>Flow m i Slave node 3 To node γ The logical link is marked as +.>Flow m i From source node β To destination node γ Is marked as->And->In the second aspect, the flow rate m i Slave node β To node 2 The logical link is marked as +.>Flow m i Slave node 2 To node 1 The logical link is marked as +.>Flow m i From source node β To destination node node 1 Is marked as->And->In a third aspect, the flow rate m i Slave node β To node 2 The logical link is marked as +.>Flow m i Slave node 2 To node η The logical link is marked as +.>Flow m i Slave node η To node 10 The logical link is marked as +.>Flow m i From source node β To destination node 10 Is marked as->And->Then there are: flow m i The transmission path from a source node to a plurality of destination nodes is denoted as traffic m i Transmission path set->And is also provided with
To describe the flow m in detail x The transmission path from a source node to a destination node is described as follows, traffic m x Slave node 10 To node 11 Transmission is thenThe logical link is noted asFlow m x Slave node 11 To node η The logical link is marked as +.>Flow m x Slave node η To node β The logical link is marked as +.>Flow m x From source node 10 To destination node β Is marked as->And->
To describe the flow m in detail y A multi-source transmission path from a source node to a plurality of destination nodes is described as follows, in a first aspect, traffic m y Slave node β To node 2 The logical link is marked asFlow m y Slave node 2 To node χ The logical link is marked as +.>Flow m y Slave node χ To node 1 The logical link is marked as +.>Flow m y From source node β To destination node 1 Is marked as->And->In the second aspect, the flow rate m y Slave node β To node 2 The logical link is marked as +.>Flow m y Slave node 2 To node γ The logical link is marked as +.>Flow m y From source node β To destination node γ Is marked as->And->In a third aspect, the flow rate m y Slave node β To node 2 The logical link is marked as +.>Flow m y Slave node 2 To node η The logical link is marked as +.>Flow m y From source node β To destination node η Is marked as->And->Then there are: flow m y Transmission paths from a source node to a plurality of destination nodes, denoted as streamsQuantity m y Transmission path set->And is also provided with
In the present invention, the subsequent link refers to the traffic m i Is set of transmission paths of (a)Middle link->There are several subsequent links, i.e.>The saidCalled flow m i In link->Is marked +.>And->
In the present invention, equivalent links refer to traffic m i Is set of transmission paths of (a)Middle link->There are multiple and links->Node with same source node 2 Links of (i.e.)>Said->Called flow m i In link->Is marked as +.>And->
In the present invention, an interference traffic means that the traffic identity that has been scheduled will be converted into a scheduled traffic. For example, flow m y After being scheduled, the flow is scheduledFlow m 1 After being scheduled, the flow is scheduled>
In the present invention, the flow rate m i On the linkThe interference flow on the node means node β Node with node 2 Inter-link-> Scheduled traffic onI.e. scheduled traffic +.>And->Is the flow rate m i In link->The interference flow set is marked as +.>And->
In the present invention, equivalent traffic means that the traffic identity that has been scheduled will be converted into scheduled traffic. For example, flow m y After being scheduled, the flow is scheduledFlow m 2 After being scheduled, the flow is scheduled>In the present invention, the flow rate m i In link->Equivalent traffic on, means that in the equivalent link setTraffic m of up-transmission i Other than scheduled trafficI.e. scheduled traffic +.>And->Is the flow rate m i In link->The equivalent flow set is marked as +.>And->
For convenience of explanation, from flow rate m i Is set of transmission paths of (a)A link is arbitrarily selected and recorded asIn the present invention, the flow rate m i In any link->The scheduling time on the table means the flow m i In link->The moment at which transmission starts is marked +.>
In the present invention, the slot s refers to a period of time spanning a certain length of time.
In the present invention, node β Node with node χ The directional transmission link between the two is denoted as l (node) β ,node χ ). The link l (node β ,node χ ) Refers to the idle time slot in l (node) β ,node χ ) During the super period of (a), there is no slot for traffic transmission. For example, time slot s is in link l (node β ,node χ ) Is of the super period of (a)Within a time slot s, and no traffic flows through l (node β ,node χ ) Then time slot s is link l (node β ,node χ ) Is allocated to the idle slot of the mobile station. When the link l (node) β ,node χ ) If a plurality of idle time slots exist, an idle time slot set is formed and marked as +.>
In the present invention, the flow rate m i On the linkThe available time slots on the base station means that +.>Time slots capable of meeting traffic scheduling constraints, denoted +.> Is->Is defined as the lower boundary value of (a). />Is->Is not included in the upper boundary value of (a).
In the present invention, the flow rate m i On the linkThe invalid time is +.>In, lead to m i Arrive at the same time as the interference flow>Is used for scheduling the time instant of (a). When the flow rate m i In link->If there are a plurality of invalid moments, a flow m is formed i In link->The invalid time set is recorded as
In the present invention, the flow rate m i On the linkThe pre-emption time slot is +.>In, schedule time according to traffic>The expected flow m i In link->And the time slot occupied thereby. When due to flow m i In link->There will be a plurality of pre-empted time slots, then traffic m is formed i On the linkThe set of pre-empted time slots is marked +.>
In the present invention, an end system refers to each of the avionics devices in the avionics system. The TSN network is composed of a plurality of end systems and switches, and physical links are interconnected between the end systems and the switches, and 8 FIFO priority queues are arranged at output ports of the end systems and the switches to provide transmission channels for messages, so that a Qbv protocol flow scheduling model of the TSN network is formed. The queue with the priority of 8 in the model is specially responsible for the transmission of ST traffic, and the other 7 priority queues are responsible for the transmission of other traffic such as AVB and BE. The TSN network controls the opening of different priority queue gates at each port through the GCL at each output port. Each element in the GCL list indicates the current slot and current door open condition, e.g., T00:1000000 indicates that at time T00 slots, the ST queue gate is open and the other queues gate is closed, as is T01:01111111 indicates that at time T01 slot, the ST queue gate is closed and the other queues gate is open. In the network topology generated by several end systems and switches as shown in fig. 2, unicast traffic is provided if there is only one destination node, and multicast traffic is provided if there is more than one destination node.
Example 1
The present embodiment is based on the PyCharm platform, and performs network scheduling design according to the fast traffic scheduling method provided by the present invention by using the network topology structure shown in fig. 2 and the ST traffic configuration information shown in table 1. Where message 2 and message 3 are multicast messages, i.e. messages having multiple destination nodes.
Table 1ST traffic configuration information
Fig. 3 shows the calculation result of the ST traffic scheduling offset in the form of a gante graph, from which a gating list of TSN output ports can be synthesized. When the time slot in the gating list is 1, consider that the least common multiple of all traffic cycles in the network is 80, and the number of entries in the gating list of each port is 80, namely from T00 to T79. The window of ST queue gate in each port, i.e. the window with a value of 10000000, is shown in table 2.
Table 2 gating list of TSN output ports
The scheduling planning result obtained by adopting the flow scheduling method of the invention can ensure that the end-to-end delay of the message is lower, as shown in figure 4, wherein the end-to-end delays of the message 1, the message 2, the message 3, the message 4 and the message 6 are all products of the corresponding message length and the hop count of the transmission path, namely the messages can realize no-waiting transmission in the network under the planning, and the transmission delay of the message 5 and the message 7 is only 2 higher than that under the condition of no-waiting transmission, and the real-time performance is still higher. In addition, analyzing the end-to-end delay of multicast message 2 on different transmission paths can also find that the end-to-end delay of traffic is closely related to the number of transmission path hops.

Claims (3)

1. A flow scheduling method of avionics system based on Qbv protocol of time sensitive network, length and cycle of the same flow under Qbv protocol of time sensitive network are fixed; flow m i Length of (2) is recorded asFlow m i The period of (2) is recorded asThe flow scheduling method is characterized by comprising the following steps:
step one, initializing flow scheduling;
step 1-1, initializing the scheduling states of all scheduling traffic;
traffic set to be schedulableFlow rate regulation of each of theStatus of degree->Initializing an unscheduled state undone;
representing a scheduling state; said->The method comprises unscheduled tunes and scheduled tunes;
the undne indicates that all traffic in the TSN network is not scheduled in the initial state, and is called unscheduled for short;
done indicates that traffic has been scheduled during the scheduling process;
step 1-2, initializing the idle time slot range of all links;
according to static routing table TAB configured in TSN network Routing And link l (node β ,node χ ) Is of the super period of (a)The free time slot set +.>Initialized to->
Step 1-3, queuing of flow scheduling;
traffic set to be schedulableSequencing from small to large according to the flow period to obtain a first sequencing set, and marking the first sequencing set as MIA;
sequencing the flows with the same period in the MIA of the first sequencing set from small to large according to the flow length to obtain a second sequencing set, and marking the second sequencing set as QM;
step 1-4, initializing the current dispatching flow;
the current dispatching flow is recorded as cur QM The method comprises the steps of carrying out a first treatment on the surface of the At initialization, cur QM Is empty;
initializing, selecting the 1st queuing flow in the second ordered set QM, and marking the 1st queuing flow as m i And subjecting said m to i As the current dispatch traffic, assign to cur QM Then there is cur QM ＝<m i >；
Step two, acquiring the current scheduling flow and a transmission path thereof;
step 2-1, obtaining current dispatching flow;
cur according to the current scheduling flow QM Obtaining traffic to be scheduled, i.e. m i ；
Step 2-2, obtaining a transmission path of the current scheduling flow;
according to static routing table TAB configured in TSN network Routing Obtaining m i Is set of transmission paths of (a)
Step 2-3, initializing the transmission link number of the current dispatching flow;
the current link is denoted as cur QP The method comprises the steps of carrying out a first treatment on the surface of the At initialization, cur QP Is empty;
initializing and selectingThe 1st path link->And putting the saidCurrent link, cur, as current scheduled traffic QP Assignment of +.>
Step 2-4, obtaining a transmission link of the current scheduling flow;
according to the current link cur QP Obtaining traffic-links to be scheduled, i.e.
Step three, calculating the upper bound and the lower bound of the available time slot range of the current dispatching flow on the current transmission link;
step 3-1, determining a calculation method of upper and lower boundaries of a time slot range;
traffic m to be scheduled i On traffic-link to be scheduledThe available time slot range is recorded asAnd-> Is thatLower boundary value of (2); />Is->Upper boundary value of (2);
judgingWhether or not it is->The first link of (a); if->Is->The first link in (i.e.)>If the source node is the end system node, executing the step 3-2; if it isNot->Step 3-3 is executed if the first link in the step is the first link;
step 3-2, calculating an upper boundary value and a lower boundary value of an available time slot range of the current scheduling flow on the first link;
will beAssignment as free time slot set->Is the left boundary of (2); calculated according to formula (1)Executing the fourth step;
is the traffic m to be scheduled i Is a period of (2); />Is the traffic m to be scheduled i Is a length of (2);
step 3-3, calculating an upper boundary value and a lower boundary value of an available time slot range of the current scheduling flow on a non-first link;
step 3-3A, findEquivalent Link set->
Step 3-3B according toStatistics m i At->Up-transmission equivalent traffic set
Step 3-3C, inFinding m in i To->The previous linkSaid->Destination node and->Is the same as the source node of (a); calculate +.>Calculate +.>
Step 3-3D, initializing equivalent flow;
utilizing an equivalent flow setEquivalent flow in (a), and iteratively updating according to the formulas (3) and (4)And->Then executing the fourth step;
represents traffic m to be scheduled i Is the number of instances of (a);
representing the number of instances of equivalent traffic; delta is clock synchronization precision;
step four, the invalid moment of the current dispatching flow on the current link is eliminated;
step 4-1, acquiring a subsequent link list of a current link aiming at the current scheduling flow;
at m i Is set of transmission paths of (a)Finding m in i In link->Subsequent link sets on
Step 4-2, obtaining an interference flow list when the current scheduling flow is transmitted on the current link;
according toStructure m i At->Interference traffic set on->
Step 4-3, determining an invalid time list of the current scheduling flow on the current link;
step 4-3A, judging the interference flow setIf not, executing the step five, and if not, executing the step 4-3B;
step 4-3B, initializing an invalid time setIs empty;
step 4-3C, utilizeIteratively updating the invalid scheduling time list according to the formula (5);
representing the number of instances of the interference traffic;
step 4-4, eliminating an invalid time list of the current scheduling flow on the current link;
updating m i At the position ofAvailable time slots ∈>And exclude invalid time sets therefrom
Step five, calculating the scheduling moment of the current scheduling flow on the current transmission link;
step 5-1, judging whether the current dispatching flow can be dispatched on the current transmission link;
according toPreliminary determination of m i At->Whether or not the upper part is schedulable; if it isThen m is i At->The step six is executed after the scheduling is not possible; if->Then m is i At->The step 5-2 is executed after the scheduling;
step 5-2, iteratively determining the scheduling time of the current scheduling flow on the current transmission link;
step 5-2A, initializing the iterative determination identifier r, and assigning r asI.e.
Step 5-2B, calculating the estimated occupied time slot of the scheduling flow under the current iteration according to the formula (7)
Step 5-2C, judgingWhether or not it belongs entirely to the free time slot set of the current link +.>If the traffic belongs to the group, the scheduling bias of the current scheduling traffic on the current transmission link can be assigned, namely +.>Let m i Scheduling state of->Adjust to done, update->And executing the step 5-2D; otherwise, executing the step 5-2G;
step 5-2D, judging whether the current scheduling flow has completed scheduling calculation on all transmission links, namelyIf the destination node is the end system, executing the step 5-2E, if not, executing the step 5-2F;
step 5-2E, updating cur QM Judging whether the scheduling calculation of all the flows is finished, if yes, executing a step seven, otherwise, executing a step 2-1;
step 5-2F, updating cur QP Executing the step 2-4;
step 5-2G, judgingIf yes, executing the step six, and if not, executing the step 5-2H;
step 5-2H, updating the iteration determination identifier r to beThe next available time slot in the time slot is returned to execute the step 5-2B;
step six, dynamically updating a flow scheduling queuing sequence;
step 6-1, determining a blocking flow which causes the current scheduling flow to be non-schedulable on the current link;
calculating an equivalent flow set according to equation (8)Any equivalent flow m y Is>
If m is y Is the smallest, then m y The identity of (a) is changed to m i Is a blocked flow of (1);
step 6-2, updating the idle transmission time slot of the related link;
finding blocked traffic m in queuing sequence QM y And subsequent traffic, compensating for the time slots occupied by these traffic on the corresponding transmission link, updating the free time slots of the associated link
Step 6-3, updating a queuing sequence of flow scheduling;
the traffic m to be scheduled in QM i Inserting blocking flow m y Before, a new queuing sequence QM is formed New type ；
Step 6-4, updating the sequencing number of the current scheduling flow;
updating cur QM Judging whether the step is finished, if so, executing the step 7, and if not, executing the step 2-1;
step seven, completing the scheduling design of all the flows;
step 7-1, generating a network scheduling Gantt chart;
converting the scheduling time of all traffic on the corresponding link into the scheduling time of the output port of the network node under the Qbv protocol of the time sensitive network, and summarizing the scheduling conditions of each output port to form a full-network scheduling Gantt chart;
step 7-2, synthesizing a gating list of the TSN output port;
and according to the actual occupation condition of each link in the scheduling Gantt chart, combining the time slot length of the gating list GCL to generate an output end gating list GCL of the node output port corresponding to each link.
2. The method for traffic scheduling of avionics systems based on the Qbv protocol of time sensitive networks according to claim 1, characterized in that: when the rapid flow scheduling based on the Qbv protocol of the time sensitive network is operated, a deterministic flow scheduling table is directly and rapidly generated in a flow dynamic sequencing and boundary equivalent conversion mode without depending on an external solver.
3. The method for traffic scheduling of avionics systems based on the Qbv protocol of time sensitive networks according to claim 1, characterized in that: the method is applied to the network topology structure generated by each aviation device in the avionics system.
CN202311375303.7A 2023-10-23 2023-10-23 Flow scheduling method of avionic system based on Qbv protocol of time sensitive network Pending CN117278487A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202311375303.7A CN117278487A (en) 2023-10-23 2023-10-23 Flow scheduling method of avionic system based on Qbv protocol of time sensitive network

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202311375303.7A CN117278487A (en) 2023-10-23 2023-10-23 Flow scheduling method of avionic system based on Qbv protocol of time sensitive network

Publications (1)

Publication Number Publication Date
CN117278487A true CN117278487A (en) 2023-12-22

Family

ID=89206132

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202311375303.7A Pending CN117278487A (en) 2023-10-23 2023-10-23 Flow scheduling method of avionic system based on Qbv protocol of time sensitive network

Country Status (1)

Country Link
CN (1) CN117278487A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117596200A (en) * 2024-01-15 2024-02-23 之江实验室 Time-sensitive network route scheduling method, electronic equipment and medium

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117596200A (en) * 2024-01-15 2024-02-23 之江实验室 Time-sensitive network route scheduling method, electronic equipment and medium
CN117596200B (en) * 2024-01-15 2024-05-07 之江实验室 Time-sensitive network route scheduling method, electronic equipment and medium

Similar Documents

Publication Publication Date Title
CN110505099B (en) Service function chain deployment method based on migration A-C learning
Bauer et al. Improving the worst-case delay analysis of an AFDX network using an optimized trajectory approach
CN112235194B (en) Method and device for scheduling delay sensitive flow on line route
Scharbarg et al. A probabilistic analysis of end-to-end delays on an AFDX avionic network
Bauer et al. Applying Trajectory approach with static priority queuing for improving the use of available AFDX resources
Prados-Garzon et al. Asynchronous time-sensitive networking for 5G backhauling
Qian et al. Qos scheduling for nocs: Strict priority queueing versus weighted round robin
CN117278487A (en) Flow scheduling method of avionic system based on Qbv protocol of time sensitive network
CN114553697B (en) Industrial wireless and TSN fusion-oriented network scheduling method
US20210073689A1 (en) Method for generating a schedule for mixed critical computer networks
Bülbül et al. SDN-based self-configuration for time-sensitive IoT networks
CN114884557B (en) Satellite time sensitive network path selection method based on network algorithm
Ginthör et al. End-to-end optimized joint scheduling of converged wireless and wired time-sensitive networks
CN114760669A (en) Routing decision method and system based on flow prediction
Hamza et al. Priority assignment on an avionics switched Ethernet Network (QoS AFDX)
Linsenmayer et al. Integration of communication networks and control systems using a slotted transmission classification model
CN111340192B (en) Network path allocation model training method, path allocation method and device
Wang et al. Deep reinforcement learning aided no-wait flow scheduling in time-sensitive networks
Yu et al. Deep reinforcement learning-based deterministic routing and scheduling for mixed-criticality flows
Li et al. Joint traffic routing and scheduling algorithm eliminating the nondeterministic interruption for TSN networks used in IIoT
Bülbül et al. Towards SDN-based dynamic path reconfiguration for time sensitive networking
Li et al. A SDN-based traffic bandwidth allocation method for time sensitive networking in avionics
CN117560106A (en) Scheduling method for time trigger flow in time sensitive network
CN115834511B (en) Gating scheduling method for periodic messages of various priorities in time-sensitive network
Singh Routing algorithms for time sensitive networks

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination