A deterministic cooperative routing and scheduling method across tsn and pon domains

Abstract

The application discloses a kind of cross TSN and PON domain deterministic cooperative routing and scheduling method, by TDM-PON is equivalent to the TSN switch (i.e. OTSN) with TAS function, the cooperative scheduling of OTSN and TSN equipment is realized, the E2E deterministic transmission problem is solved.And the TW constraint formula for OTSN and TSN joint scheduling and the cross-domain cooperative routing and scheduling scheme based on the greedy algorithm are proposed to plan the position of TW, meet the delay, jitter transmission demand of various businesses, compared with traditional non-cooperative scheduling scheme, the schedulability is improved, it is a kind of E2E deterministic transmission capacity, meet the routing and scheduling scheme of the demand of multiple types of businesses.

Description

Technical Field

[0001] This invention relates to the field of industrial internet communication technology, and in particular to a deterministic cooperative routing and scheduling method across TSN and PON domains. Background Technology

[0002] With the advancement of the Industrial Internet, the convergence of information technology (IT) and operational technology (OT) has become an inevitable trend. Technologies such as industrial virtual reality (VR) / augmented reality (AR) will rapidly become widespread, helping engineers remotely monitor equipment status in factories in real time. Furthermore, to expand the scope of machine collaboration, programmable logic controllers (PLCs) can be deployed at edge data centers to achieve centralized control. The interaction of control information between PLCs located in the data center and field devices requires the support of an access network. Time-sensitive (TS) industrial applications, such as controller-to-controller and industrial robots, have extremely high latency (100μs-2ms) and jitter (1μs-2ms) requirements, demanding that the access network possess deterministic transmission capabilities, ensuring that latency and jitter have upper bounds and meet the transmission requirements of the business.

[0003] Time-Sensitive Networking (TSN) plays a crucial role in the Industrial Internet. The TSN working group has proposed a series of mechanisms, such as frame preemption and Time-Aware Shaping (TAS), to ensure low latency and jitter for TS industrial services. TAS technology, in particular, can control the transmission state of the transmission gating queue and set up a dedicated transmission window (TW) for TS services to avoid interference from other services, thereby achieving sub-millisecond deterministic latency and zero jitter. However, in large-scale network scenarios, this may require a large number of TSN switches, which not only increases costs but also makes the configuration of TSN switches (the calculation of the gating list) extremely complex.

[0004] Time-division multiplexing passive optical networks (TDM-PON), as a point-to-multipoint (P2MP) network technology, have attracted the attention of researchers. In industrial automation, a production line typically includes a line controller (L-PLC) and multiple machines working together to complete a task. Each machine usually also includes a dedicated PLC and multiple field devices (sensors / actuators). Here, the PLC (or L-PLC) can be deployed in the same geographical location as the field devices (hereinafter referred to as field PLC), or it can be deployed at an edge data center (DC) (hereinafter referred to as DC-PLC), which can be determined by the manufacturer. Figure 1As shown, field devices and PLCs are connected via TSN devices and then connected to TDM-PON, thereby connecting to the DC-PLC. Industrial automation based on TDM-PON typically includes the following three communication types: controller-to-controller (C2C / L2C) communication, controller-to-field device (C2D) communication, and field device-to-computer (D2Cmp) communication.

[0005] 1. Controller-to-Controller (C2C / L2C) Communication

[0006] In industrial production lines, multiple machines often need to work together to perform tasks. Therefore, L-PLCs are typically used to monitor and control the entire production line to ensure normal production. For example... Figure 1 As shown, the yellow stream represents C2C communication. This type of communication typically includes the following:

[0007] 1) Between the two ONUs connected to the PLC in a PON (within the PON): such as Figure 1 As shown in (a), the two PLCs are connected to different ONUs inside the PON.

[0008] 2) Between PLCs connected to ONUs of two PONs (across PONs): such as Figure 1 As shown in (b), the two PLCs are connected to the ONUs in the two PONs respectively.

[0009] 3) Between a PON-linked PLC and a DC-PLC (via PON): The PLC is connected to the ONU, and the DC-PLC is connected to the OLT within the same PON (e.g., ...). Figure 1 (a) or another PON (such as Figure 1 (b)

[0010] 2. Controller-to-Field Device (C2D) Communication

[0011] A machine typically consists of a controller (PLC) and a series of field devices (i.e., sensors / actuators). For example... Figure 1 As shown, the red stream represents C2D communication. This type of communication typically includes the following:

[0012] 1) Between the PLC and field devices connected to the two ONUs of the PON (within the PON): such as Figure 1 As shown in (a), the PLC and field devices are connected to different ONUs inside the PON.

[0013] 2) Between PON-connected field devices and DC-PLC (via PON): The field device is connected to the ONU, and the DC-PLC is connected to the OLT within the same PON (e.g., Figure 1 (a) or another PON (such as Figure 2 (b)

[0014] 3. Communication between field devices and the computer (D2Cmp)

[0015] Operators can collect status information from field equipment using computers in the data center, thereby understanding the operational status of the field equipment and other related information. For example... Figure 1 As shown, the green stream represents D2Cmp communication. This type of communication typically includes the following:

[0016] 1) PON-connected field devices and computers (via PON): Field devices are connected to the ONU, while the computer is typically located at the edge DC, and the DC is connected to the OLT within the same PON (e.g., ...). Figure 1 As shown in (a)).

[0017] Regarding the communication requirements of Industrial Internet services, Table 1 summarizes the corresponding flow types and connection networks for the three communication types mentioned above. C2C / L2C and C2D communication types include isochronous flows, cyclic flows, event flows, and configuration and diagnostic flows. D2Cmp communication types include event flows, configuration and diagnostic flows, and audio / video flows. The flow types in Table 1 can be roughly divided into three categories: 1) isochronous flows, 2) cyclic flows, and 3) burst flows (including event, configuration and diagnostic, and audio / video flows).

[0018] Table 1 Summary of Industrial Internet Flow Types

[0019]

[0020] Table 2 lists the period, delay, and jitter requirements corresponding to the three flow types mentioned above. Isochronous and cyclic flows typically have strict requirements for delay and jitter, with isochronous flows having the most stringent jitter requirement (1μs), while cyclic flows have relatively more lenient jitter requirements (200μs-2ms). Burst flows have more lenient delay and jitter requirements than isochronous and cyclic flows.

[0021] Table 2 Demand for Three Types of Flows

[0022]

[0023]

[0024] In summary, industrial data flows have extremely stringent end-to-end latency and jitter requirements, necessitating the access network to possess deterministic transmission capabilities.

[0025] Existing heterogeneous network solutions employ non-joint scheduling schemes, where the TSN and PON domains each guarantee deterministic transmission within their respective domains. In the TSN domain, TAS technology is typically used, leveraging Satisfiability Model Theory (SMT) and Integer Linear Programming (ILP) to plan the transit-time (TW) positions of flows to ensure deterministic transmission. In the PON domain, schemes include Fixed Bandwidth Allocation (FBA), Dynamic Bandwidth Allocation (DBA), and Time-Aware Deterministic Bandwidth Allocation (TA-DetBA).

[0026] Among them, TAS technology is one of the technologies that has attracted the most attention from researchers in TSN. For example Figure 2 As shown, different priority queues are controlled by different transmission gates. When a gate is open, data packets within the queue are allowed to be transmitted; when the gate is closed, transmission is not allowed. The transmission gates of all queues are controlled by a gating list, which contains the opening and closing information for each gate in each time slot. In TAS technology, high-priority services (TS streams) are isolated by opening and closing these gates to prevent interference from other priority services, thus ensuring deterministic latency and jitter. In FBA, the OLT periodically allocates a fixed amount of TW (i.e., bandwidth) to the ONU, regardless of whether the ONU requires it. Even if the TW allocated to an ONU is unused, it will not be reassigned to other ONUs. In GPON, the size of the TW is related to the peak rate of the stream, while the TW period is related to the uplink frame duration (125μs). In DBA, the OLT and ONU rely on a "request and grant" mechanism to allocate TW. Each ONU reports the traffic load status in its queue after a polling cycle. Once the OLT receives the traffic loads from all ONUs, it allocates a TW size and start time (offset in the next uplink frame) to each ONU according to a bandwidth allocation algorithm. The TW size allocated by the OLT to the ONU is related to the ONU's traffic load, while the TW start time is allocated sequentially according to a certain order (e.g., priority). In TA-DetBA, the OLT allocates the TW size based on the flow's bandwidth requirements, and simultaneously constrains the TW position of the flow based on requirements such as the flow's time of arrival (AT), period, latency, and jitter. This position allocation is performed within the time-over-period of all flows, thus satisfying the deterministic requirements of industrial flow latency and jitter. Figure 3 As shown, the TW location is restricted to within AT and the upper delay bound (UBD), which ensures deterministic service latency. Meanwhile, the offset between the start time and the arrival time of each TW represents the latency of each data packet, and the maximum difference between these offsets is the jitter. By limiting the magnitude of the offset difference, deterministic jitter of the flow can be guaranteed.

[0027] Existing heterogeneous network solutions employ non-cooperative scheduling, where the TSN and PON domains each guarantee deterministic transmission within their respective domains. In the TSN domain, SMT and ILP can be used to plan transmission flow (TW) to ensure determinism. However, in the PON domain, traditional FBA and DBA schemes struggle to guarantee deterministic transmission. In traditional FBA, the OLT periodically allocates TW to ONUs. To simplify configuration, the TW period is typically a multiple or divisor of the uplink frame duration (125μs). However, industrial flow periods are diverse, potentially leading to a mismatch between the TW period and the industrial flow period. For example... Figure 3 As shown, the TW period of the FBA is 62.5μs, which does not match the flow period (100μs), thus causing variations in the interval between industrial flow data packets (e.g., T1 and T2), resulting in jitter. This is unacceptable for some industrial flows (e.g., isochronous flows). In traditional DBA, within each DBA cycle, the OLT allocates an appropriate TW size to each ONU based on the traffic load reported by the ONU. However, the OLT only guarantees that the bandwidth requirements of the industrial flow (i.e., the TW size) are met, but does not guarantee the TW position. Figure 3 As shown, the DBA allocates a TW (Transmission Wave) to the ONU in each uplink frame; however, the location of the TW is uncertain, leading to uncertainty in service latency and jitter. TA-DetBA is a deterministic bandwidth allocation scheme for TDM-PON, which can achieve end-to-end deterministic transmission through cooperation with the TSN domain. However, non-cooperative scheduling schemes have certain problems: on the one hand, due to the lack of industrial flow parameters and TW scheduling information from different domains, multiple domains cannot coordinate scheduling, which may cause congestion when industrial flows from multiple TSN domains converge to the PON domain, thus reducing service schedulability; on the other hand, since the TSN domain and PON domain cannot be jointly scheduled, delay budget allocation needs to be done in advance, and this inflexible allocation method may also lead to a decrease in schedulability.

[0028] In summary, while TDM-PON is simple to configure and is a mature, high-bandwidth, low-cost technology resistant to strong electromagnetic interference, it makes it possible to replace some TSN nodes and thus simplify network configuration. However, this brings a new challenge: how to achieve deterministic transmission and global optimization of end-to-end (E2E) TS industrial flows across TSN and PON domains. Summary of the Invention

[0029] The purpose of this invention is to propose a deterministic cooperative routing and scheduling method across TSN and PON domains. It equates TDM-PON to a TSN equivalent model with TAS functionality, namely "OTSN", and performs cross-domain joint routing and scheduling together with traditional TSN switches. In this way, the latency budgets of PON domain and TSN domain are treated as a whole to achieve global optimization of E2E TS industrial services and improve the schedulability of services across the entire network.

[0030] To achieve the above objectives, the present invention provides the following technical solution:

[0031] A deterministic cooperative routing and scheduling method across TSN and PON domains includes the following steps:

[0032] S1. Collect parameters of all TS streams, including the source node, destination node, start time, period, latency, jitter, and data size of the TS stream;

[0033] S2. Equivalent TDM-PON to an OTSN switch;

[0034] S3. Calculate the K shortest paths for all TS flows;

[0035] S4. Sort all TS flows in ascending order of the minimum maximum tolerable waiting time among their K paths;

[0036] S5. Schedule all TS streams in sequence. During scheduling, each stream will have a cost value (COST) of α × ∑ [u，v]∈L N [u，v] Minimal scheduling scheme, N [u，v] This refers to the number of time slot fragments on the link from node u to node v. A time slot fragment is a consecutive time slot that is larger than the TW protection width and smaller than the minimum packet size of all flows.

[0037] Furthermore, the specific implementation process of step S2, which equates TDM-PON to an OTSN switch, is as follows: TDM-PON is regarded as a whole, and the ports of ONU and OLT are regarded as the access port and aggregation port of the OTSN switch, respectively; the flow mapping module of OTSN is equivalent to the flow mapping module of ONU and a port ID is added; the transmission container of ONU is equivalent to the T queue in the OTSN switch, and there is a physical guard band between the TW of the two queues in OTSN; ODN is equivalent to the delay gate located in front of the T queue, and the start time of TW in OTSN is equivalent to the time when TW arrives at OLT in TDM-PON.

[0038] Furthermore, the method for sorting all TS streams in step S4 is as follows:

[0039] First, calculate the maximum tolerable waiting time on the K shortest paths of each flow, and the maximum tolerable waiting time on path r. for:

[0040]

[0041] Where F is the set of industrial flows, K i Let D be the set of K shortest paths for flow i, where [u, v] is the link between node u and node v. i It is the end-to-end latency requirement of streaming i. It is the propagation delay between nodes u and v. W is the processing latency within node v. i τ is the data size of stream i, and τ is the maximum time synchronization error within the network.

[0042] Then, calculate the minimum maximum tolerable waiting delay among the K paths of TS flow i. Then sort them in ascending order.

[0043] Furthermore, the method for scheduling all TS streams sequentially in step S5 is as follows:

[0044] a) Select a flow in sequence according to the order of the flows, and schedule it sequentially on its K paths;

[0045] b) Select one path from the K paths, and allocate the TW start time for each node from the source node to the destination node. The allocation of the start time starts from the first time slot, traverses all time slots, and checks whether the start time of TW meets the constraints. Select the time slot with the lowest COST value in the path. If all constraints are met, it means that the flow is successfully scheduled at this node, and then continue to schedule the next node. If all nodes are successfully scheduled, it means that the flow is successfully scheduled on this path.

[0046] c) Determine whether the flow was successfully scheduled on path r. If the scheduling was successful, calculate the COST value of the path and record the path, COST value, and TW start time of each node in the path. After recording, continue to schedule the next path. If the scheduling was unsuccessful, continue to schedule the next path.

[0047] d) Determine if there are any unscheduled paths. If they exist, continue scheduling; otherwise, determine if there are any successfully scheduled paths.

[0048] e) Determine if a successfully scheduled path exists. If it does, select the path with the lowest COST and configure it according to the TW start time of each node in the scheduling scheme. If it does not exist, select the next flow for scheduling.

[0049] f) Determine if there are any unscheduled flows. If so, continue scheduling the next flow; otherwise, complete the scheduling of all flows and configure the actual network based on the scheduling results of all flows.

[0050] Furthermore, the constraints in step b) include:

[0051] 1) Range constraints:

[0052] TW start time of flow i on the link [u, v] from node u to v The following constraints should be met:

[0053] i = F; u, v ∈ V

[0054]

[0055] Range constraint means: Suppose flow i runs from node u to node v, the start time of the first period TW of flow i should be greater than 0 and less than its period P. i W i It is the data size of stream i;

[0056] 2) Stream transmission constraints:

[0057] TW start time of flow i on the link [v, w] from node v to w The following constraints should be met:

[0058] i∈F; u, v, w∈V

[0059]

[0060] The stream transmission constraint states: Suppose stream i flows from node u to node v and then to node w. The start time of stream i at the current node v must be greater than the sum of the start time of stream i at the previous node u, the propagation delay between the two nodes, the processing delay of node v, the time synchronization error, and the sending delay of stream i. If node u is the source node B of the stream... i ,but A i Let i be the time when stream i is sent at the source node; It is the propagation delay between nodes u and v. τ is the processing latency within node v, and τ is the maximum time synchronization error within the network.

[0061] 3) End-to-end delay constraints:

[0062] The end-to-end delay constraint should satisfy the following constraints:

[0063] i∈F; u,E i ∈V

[0064]

[0065] End-to-end delay constraint representation: Let flow i originate from source node B i After passing through multiple nodes to the destination node E i The time when flow i reaches the destination node and the start time A of the flow. i The difference between them must be greater than the end-to-end delay D of stream i. i ; It is the processing latency within node u;

[0066] 4) Isolation constraints:

[0067] The start time of TW for flows i and j on the link from node u to node v. The following constraints should be satisfied if u is a TSN node:

[0068] i, j ∈ F; u, v ∈ V

[0069]

[0070] If u is an OTSN node:

[0071]

[0072] The isolation constraint means that if both flows i and j run from node u to node v, the TW between flows i and j cannot overlap; G is the size of the protection bandwidth.

[0073] Furthermore, in step f), during actual network configuration, the start time of the ONU will be calculated based on the start time of the flow at the OTSN. The start time of the TW of flow i at the OTSN node u is... Then the actual start time of the ONU for:

[0074]

[0075] Among them, E i It is the destination node of stream i. P is the processing latency within node u. i It is the period of flow i.

[0076] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0077] This invention proposes a deterministic cooperative routing and scheduling method across TSN and PON domains. By equating TDM-PON with a TAS-enabled TSN switch (i.e., OTSN), it achieves cooperative scheduling of OTSN and TSN devices, solving the E2E deterministic transmission problem. Furthermore, it proposes a TW constraint formula for joint scheduling of OTSN and TSN and a cross-domain cooperative routing and scheduling scheme based on a greedy algorithm to plan the location of TWs and meet the latency and jitter transmission requirements of various services. Compared with traditional non-cooperative scheduling schemes, it improves schedulability and is a routing and scheduling scheme with E2E deterministic transmission capability that meets the needs of multiple types of services. Attached Figure Description

[0078] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0079] Figure 1 This is an industrial internet communication type based on TDM-PON.

[0080] Figure 2 This is a schematic diagram of TAS technology.

[0081] Figure 3 This is a schematic diagram of the TA-DetBA scheme.

[0082] Figure 4 This is a schematic diagram illustrating the problems with traditional bandwidth allocation schemes.

[0083] Figure 5 This is a diagram of the TSN equivalent model of the TDM-PON proposed in this invention.

[0084] Figure 6 This is a flowchart of the deterministic cooperative routing and scheduling method across TSN and PON domains proposed in this invention. Detailed Implementation

[0085] To better understand this technical solution, the method of the present invention will be described in detail below with reference to the accompanying drawings.

[0086] 1. Network and service model

[0087] This invention provides a deterministic cooperative routing and scheduling method across TSN and PON domains, based on a network architecture where network nodes include multiple field devices, multiple TSN switches, multiple ONUs, one OLT, and one DC-PLC. In the network, field devices are connected by TSN switches, which are interconnected in a mesh configuration. The network is divided into multiple TSN domains, each with several directly or indirectly connected TSN switches, and different TSN domains may also be interconnected. ONUs and OLT nodes are considered a TDM-PON domain; in the following text, TDM-PON is equivalent to an OTSN switch, and therefore, TDM-PON can be considered a network node. Each TSN domain may have one or more TSN switches connected to ONUs, each TSN switch connects to one ONU, and each ONU connects to one or more TSN switches. The OLT is directly connected to the DC-PLC, and control of the field devices is performed at the DC-PLC. This patent only considers transmission from the field device to the DC-PLC (uplink). The following are the definitions of network parameters and service parameters:

[0088] 1) First, define the network parameters as follows:

[0089] • The set of network nodes: V

[0090] • The set of links between two nodes: L

[0091] The link between node u and node v: [u, v]

[0092] • Propagation delay between nodes u and v:

[0093] • Processing latency within node v (if the node is an OTSN, then it is the processing latency of the ONU + the processing latency of the OLT + the propagation latency from the ONU to the OLT):

[0094] • Protected bandwidth within “OTSN”: G

[0095] • Maximum time synchronization error within the network: τ

[0096] 2) Then, the parameters for the time-delay sensitive flow are defined. The parameters for the industrial flow are as follows:

[0097] • Collection of industrial flows: F

[0098] • Source node of stream i: B i , i∈F

[0099] • Destination node of stream i: E i , i∈F

[0100] The set of K shortest paths for flow i: K i , i∈F

[0101] • End-to-end jitter requirements for streaming: J i , i∈F

[0102] • Streaming end-to-end latency requirements: D i , i∈F

[0103] • Maximum tolerable waiting time for flow i on path r:

[0104] • The time when stream i is sent at the source node: A i , i∈F

[0105] • Period of flow i: P i , i∈F

[0106] • Data size of stream i: W i , i∈F

[0107] Decision variables:

[0108] • Window start time on the link from node u to node v for flow i:

[0109] • Number of time slot fragments on the link from node u to node v: N [u，v]

[0110] All time and bandwidth-related parameters are expressed in time slot form.

[0111] 2 Transmission Window Constraints

[0112] 1) Range constraints:

[0113] Suppose flow i runs from node u to node v. The range constraint stipulates that the start time of the first period TW of flow i should be greater than 0 and less than its period (P). i Therefore, the TW start time of flow i on the link [u, v] from node u to v. The following constraints should be met:

[0114] i = F; u, v ∈ V

[0115]

[0116] 2) Stream transmission constraints:

[0117] Suppose stream i flows from node u to node v and then to node w. Stream transmission constraints stipulate that the start time of stream i at the current node v must be greater than the sum of the start time of stream i at the previous node u, the propagation delay between the two nodes, the processing delay at node v, the time synchronization error, and the transmission delay of stream i. Therefore, the TW start time of stream i on the link [v, w] from node v to w is... The following constraints should be met:

[0118] i∈F; u, v, w∈V

[0119]

[0120] If node u is the source node B of the flow i ,but

[0121] 3) End-to-end delay constraints:

[0122] Let flow i originate from source node B. i After passing through multiple nodes to the destination node E i The end-to-end delay constraint specifies the time it takes for stream i to reach the destination node and the start time A of the stream. i The difference between them must be greater than the end-to-end delay D of stream i. i Therefore, the following constraints should be satisfied:

[0123] i∈F; u,E i ∈V

[0124]

[0125] 4) Isolation constraints:

[0126] Suppose flows i and j both originate from node u to node v, and isolation constraints stipulate that the transit times (TWs) of flows i and j cannot overlap. Therefore, the start times of the TWs for flows i and j on the link from node u to node v are... The following constraints should be satisfied if u is a TSN node:

[0127] i, j ∈ F; u, v ∈ V

[0128]

[0129]

[0130] If u is an OTSN node:

[0131]

[0132] The cross-TSN and PON domain cooperative routing and scheduling method proposed in this invention aims to optimize the schedulability of flows. The process is as follows: Figure 1As shown, the process includes the following steps: S1, collecting parameters for all TS streams; S2, equating TDM-PON with an OTSN switch; S3, calculating the K shortest paths for all TS streams; S4, sorting all TS streams; S5, scheduling all TS streams in sequence. A detailed description follows.

[0133] S1. Collect parameters from all TS streams.

[0134] The parameters of the industrial stream are collected through the interface between the OLT and external devices (such as the Collaborative Transmission Interface, CTI). These parameters include the source node, destination node, start time, period, latency, jitter, and data size of the TS stream.

[0135] S2. Equivalent TDM-PON to OTSN switch.

[0136] TDM-PON is equivalent to an "OTSN" switch, used to achieve joint scheduling of the PON domain and TSN domain. Figure 5 This section demonstrates TDM-PON and its TSN equivalent model (OTSN). TDM-PON is viewed as a single entity, where TS flows enter ONU ports and exit through OLT ports. Therefore, the ports of the ONU and OLT can be considered as the access ports and aggregation ports of the OTSN switch, respectively. After entering the ONU port, the TS flow is mapped (based on VLAN ID or priority) to different transport containers (T-CONTs) by the flow mapping module. Thus, the OTSN's flow mapping module is equivalent to the ONU's flow mapping module, but it adds a port ID to distinguish flows accessing different ONUs. In TDM-PON, the OLT assigns a TW to each T-CONT, so the T-CONT is equivalent to a "T queue" in the OTSN switch. T-CONT transmission is controlled by TA-DetBA, and the control method of TA-DetBA is similar to the gating mechanism of TAS; therefore, TA-DetBA can be equated to the gating mechanism of TAS. Traditional TSN switches with TAS mechanisms allow two TWs to overlap, but in OTSNs, a physical guard band (G) must exist between the TWs in the two queues to avoid collisions between TWs during uplink transmission. Data packets in the T-CONT are aggregated to the OLT via the Optical Distribution Network (ODN), introducing propagation delay. The ODN can be equated as a delay gate located before the T queue, simulating the different propagation delays of the ONU. It's important to note that because data packets experience a fixed delay before entering the T queue, the start time of a TW in the OTSN is the same as the time it takes for a TW to arrive at the OLT in TDM-PON; therefore, the start time of a TW in the ONU needs to be recalculated based on this time.

[0137] S3. Calculate the K shortest paths for all TS flows.

[0138] There are two forms of flow isolation: time slot isolation and path isolation. By selecting different routes, conflicts can be avoided and schedulability can be improved. Therefore, before scheduling, the K shortest paths for all TS flows must first be calculated (e.g., using KSP). In subsequent scheduling, appropriate paths can be selected from the K paths for scheduling.

[0139] S4. Sort all TS streams.

[0140] To schedule as many flows as possible, the TS flows need to be sorted. In this patent, they are sorted in ascending order based on the minimum maximum tolerable waiting time across each flow's K shortest paths. First, the maximum tolerable waiting time on each flow's K shortest paths is calculated. The maximum tolerable waiting time on path r... for:

[0141] i∈F;r∈K

[0142]

[0143] Then, calculate the minimum maximum tolerable waiting delay among the K paths of TS flow i. Then sort them in ascending order.

[0144] S5. Schedule all TS streams in sequence.

[0145] like Figure 6 As shown, after all TS streams are sorted, they will be scheduled sequentially. Each stream enters a different queue to ensure that packets between streams do not overlap. During scheduling, each stream will select a queue with a cost (COST = α × ∑). [u，v]∈L N [u，v] The scheduling scheme with the minimum value of N. [u，v] This represents the number of time slot fragments on the link from node u to node v. A time slot fragment refers to a consecutive time slot that is larger than the inter-Wave protection bandwidth but smaller than the minimum packet size for all flows. Reducing time slot fragments aims to improve flow schedulability.

[0146] Specifically, step S5 involves scheduling all TS streams sequentially as follows:

[0147] a) Select a flow in sequence according to the order of the flows, and schedule it sequentially on its K paths;

[0148] b) Select one path from the K paths and allocate the TW start time for each node from the source node to the destination node. The allocation of the start time starts from the first time slot, traverses all time slots, and determines whether the start time of TW satisfies all the above constraints (range constraint, stream transmission constraint, end-to-end delay constraint, isolation constraint). Select the time slot with the lowest COST value in the path. If all constraints are satisfied, it means that the stream is successfully scheduled at this node, and then continue to the next node for scheduling. If all nodes are successfully scheduled, it means that the stream is successfully scheduled on this path.

[0149] c) Determine whether the flow was successfully scheduled on path r. If the scheduling was successful, calculate the COST value of the path and record the path, COST value, and TW start time of each node in the path. After recording, continue to schedule the next path. If the scheduling was unsuccessful, continue to schedule the next path.

[0150] d) Determine if there are any unscheduled paths. If they exist, continue scheduling; otherwise, determine if there are any successfully scheduled paths.

[0151] e) Determine if a successfully scheduled path exists. If it does, select the path with the lowest COST and configure it according to the TW start time of each node in the scheduling scheme. If it does not exist, select the next flow for scheduling.

[0152] f) Determine if there are any unscheduled flows. If so, continue scheduling the next flow; otherwise, complete the scheduling of all flows and configure the actual network based on the scheduling results of all flows.

[0153] In actual network configuration, the start time of the ONU will be calculated based on the start time of the flow at the OTSN. The start time of flow i at the TW of OTSN node u is... Then the actual start time of the ONU for:

[0154]

[0155] Among them, E i It is the destination node of stream i. P is the processing latency within node u. i It is the period of flow i.

[0156] This invention equates TDM-PON to a TSN switch (i.e., OTSN) with TAS functionality, achieving collaborative scheduling of OTSN and TSN devices and solving the end-to-end deterministic transmission problem. Furthermore, it proposes a TW constraint formula for joint scheduling of OTSN and TSN, and a cross-domain collaborative routing and scheduling scheme based on a greedy algorithm to plan the location of TWs, meeting the latency and jitter transmission requirements of various services. Compared with traditional non-cooperative scheduling schemes, it improves schedulability and is a routing and scheduling scheme with end-to-end deterministic transmission capability that meets the needs of multiple service types.

[0157] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. However, these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A deterministic cooperative routing and scheduling method across TSN and PON domains, characterized in that, Includes the following steps: S1. Collect parameters of all TS streams, including the source node, destination node, start time, period, latency, jitter, and data size of the TS stream; S2. Equivalent TDM-PON to an OTSN switch; the specific implementation process is as follows: treat TDM-PON as a whole, and regard the ports of ONU and OLT as the access port and aggregation port of the OTSN switch, respectively; regard the flow mapping module of OTSN as the flow mapping module of ONU and add port ID; regard the transmission container of ONU as the T queue in OTSN switch, and there is a physical guard band between the TW of the two queues in OTSN; regard ODN as the delay gate located in front of the T queue, and regard the start time of TW in OTSN as the time of TW arrival at OLT in TDM-PON; S3. Calculate the K shortest paths for all TS flows; S4. Sort all TS flows in ascending order of the minimum maximum tolerable waiting delay among their K paths; the method for sorting all TS flows is as follows: First, calculate the maximum tolerable waiting time on the K shortest paths for each flow. Maximum tolerable waiting time for: , , in, It is a collection of industrial flows. It is a flow The set of K shortest paths, For nodes and nodes The link between them It is a flow End-to-end latency requirements It is a node and The inter-propagation delay, It is a node Internal processing latency, It is a flow Data size, It is the maximum time synchronization error within the network; Then, calculate the TS stream. The minimum maximum tolerable waiting time among the K paths And sort them in ascending order; S5. Schedule all TS streams sequentially, selecting a bit value for each stream during scheduling. Minimal scheduling scheme, For nodes To the node The number of time slot fragments on the link. A time slot fragment refers to a consecutive time slot that is larger than the TW protection width and smaller than the minimum packet size of all flows.

2. The deterministic cooperative routing and scheduling method across TSN and PON domains according to claim 1, characterized in that, Step S5 involves scheduling all TS streams sequentially as follows: a) Select a flow in sequence according to the order of the flows, and schedule it sequentially on its K paths; b) Select one path from the K paths, and allocate the TW start time for each node from the source node to the destination node. The allocation of the start time starts from the first time slot, traverses all time slots, and checks whether the start time of TW meets the constraints. Select the time slot with the lowest COST value in the path. If all constraints are met, it means that the flow is successfully scheduled at this node, and then continue to schedule the next node. If all nodes are successfully scheduled, it means that the flow is successfully scheduled on this path. c) Determine if the diversion is on the path If the scheduling is successful, calculate the COST value of the path and record the path, COST value, and TW start time of each node in the path. After recording, continue to schedule the next path; if the scheduling is unsuccessful, continue to schedule the next path. d) Determine if there are any unscheduled paths. If they exist, continue scheduling; otherwise, determine if there are any successfully scheduled paths. e) Determine if a successfully scheduled path exists. If it does, select the path with the lowest COST and configure it according to the TW start time of each node in the scheduling scheme. If it does not exist, select the next flow for scheduling. f) Determine if there are any unscheduled flows. If so, continue scheduling the next flow; otherwise, complete the scheduling of all flows and configure the actual network based on the scheduling results of all flows.

3. The deterministic cooperative routing and scheduling method across TSN and PON domains according to claim 2, characterized in that, The constraints in step b) include: 1) Range constraints: flow At the node arrive Link TW start time The following constraints must be met: , (1), Range constraint representation: Let the flow From node To the node ,flow The start time of the first cycle of TW should be greater than 0 and less than its cycle. ; It is a flow Data size; 2) Stream transmission constraints: flow At the node arrive Link TW start time The following constraints must be met: , (2), Stream transmission constraint representation: Let the stream... From node To the node Then to the node ,flow At the current node The start time must be greater than the flow The previous node Start time, propagation delay between two nodes, Node processing latency, time synchronization error, and stream The sum of the transmission delays; if the node It is the source node of the stream. ,but , For flow At the time of transmission from the source node; It is a node and The inter-propagation delay, It is a node Internal processing latency, It is the maximum time synchronization error within the network; 3) End-to-end delay constraints: The end-to-end delay constraint should satisfy the following constraints: , (3) , End-to-end delay constraint representation: Let the flow From the source node via multiple nodes to the destination node ,flow Time to reach the destination node and time to start the stream The difference between them is greater than the flow. end-to-end delay ; It is a node Internal processing latency; 4) Isolation constraints: flow and At the node To the node Start time of TW on the link The following constraints should be satisfied if For TSN nodes: , , (4), if For OTSN nodes: , (5) , Isolation constraint representation: Let the flow and All from nodes To the node ,flow and The TW characters cannot overlap; It refers to the size of the protection bandwidth.

4. The deterministic cooperative routing and scheduling method across TSN and PON domains according to claim 2, characterized in that, Step f) In actual network configuration, the ONU's start time will be calculated based on the flow's start time at the OTSN. At OTSN node The start time of Taiwan is Then the actual start time of the ONU for: , in, It is a flow The destination node It is a node Internal processing latency, It is a flow The cycle.

Images