A fast rerouting method with backtracking mechanism and QoS strategy

Abstract

The invention discloses a fast rerouting method with a backtracking mechanism and a QoS strategy, and belongs to the field of satellite network communication technology. The method first configures the main and backup routing tables for each node in the satellite network, and finds the corresponding routing entry set from the destination network segment; when the node receives the message, it records the incoming interface number and the source MAC address of the message, and parses the destination IP address in the message; after receiving the message, the node will search the routing table according to the destination IP address, obtain the main and backup routing entries and the corresponding outgoing interface status of the routing entries, and check in turn whether the routing entries can be used for message forwarding; if the message forwarding cannot be achieved based on the existing main and backup routing entries, the message is backtracked, and the message is backtracked to the upstream node or discarded based on the QoS strategy; finally, the upstream node re-routes the backtracked message. The method can effectively solve the problem of frequent multi-link failures in inter-satellite links and satellite-to-ground links in a satellite network environment.

Description

Technical Field

[0001] The invention belongs to the technical field of satellite network communication, and in particular relates to a fast rerouting method with a backtracking mechanism and a QoS (Quality of Service) strategy. Background Art

[0002] Due to the influence of various factors such as cosmic radiation, solar activity and electromagnetic interference, the links in the satellite network are prone to frequent interruptions. Such unpredictable interruptions pose a severe challenge to the stability and reliability of the network. At the same time, when the satellite nodes or links in the network fail or become congested, the nodes or links will fail, causing the network transmission performance to drop sharply. Therefore, the development of network communication technologies with high adaptability and fault recovery capabilities is crucial to ensure the transmission stability and reliability of satellite networks under extreme conditions.

[0003] Traditional control plane-based rerouting technology usually relies on routing reconvergence to solve link failure problems, but this process is often time-consuming, which constitutes an obstacle to providing stable services. Compared with the control plane-based rerouting method, the forwarding plane fast rerouting (FRR) has obvious advantages of fast fault handling response and low data packet loss rate, which is particularly suitable for dealing with unforeseen link interruptions.

[0004] When unpredictable interruptions or failures occur, the fast rerouting method shows its important value in ensuring the stability and high reliability of satellite networks. FRR uses pre-calculated backup paths or set forwarding rules to quickly switch routes when the main path fails, thereby ensuring the complete forwarding of messages.

[0005] However, when the primary path fails and switches to the backup or equal-cost path, if these protection links also fail, especially when both the primary and backup forwarding paths fail, the existing FRR solution can only choose to discard the message; at the same time, FRR technology is usually only applicable to situations where the link load is light when solving the link failure problem. When the link load is heavy, transferring traffic to other paths may increase the burden on these links, which may eventually lead to link overload and congestion.

[0006] In general, although FRR technology provides important guarantees for the stability and reliability of satellite network communications, link failure recovery under extreme conditions and traffic management under high load conditions are still challenges that need further research and resolution. Summary of the invention

[0007] Purpose of the invention: In view of the problems existing in the above-mentioned FRR method, the present invention discloses a fast rerouting method with a backtracking mechanism and a QoS (Quality of Service) strategy. The method solves the problem of multi-link failure by backtracking, and performs message backtracking and forwarding based on the QoS strategy for different service types and link bandwidth states, thereby significantly reducing the packet loss rate of the entire satellite network and improving the reliability of satellite network communication.

[0008] To achieve the above object, the technical solution adopted by the present invention is: a fast rerouting method with a backtracking mechanism and a QoS strategy, the method comprising the following specific steps:

[0009] S1. Preprocessing: Each node in the satellite network is configured with a primary and backup routing table. The corresponding routing entry set is found by the destination network segment. Each routing entry should contain the next hop node interface IP address and the outbound interface number, and the interface status table is searched to obtain the outbound interface status.

[0010] S2, message reception: When a node receives a message, it records the incoming interface number and source MAC address of the message, and parses the destination IP address in the message;

[0011] S3, message forwarding: After receiving the message, the node will search the routing table according to the parsed destination IP address, obtain the primary and backup routing entries and the status of the corresponding outbound interfaces of the routing entries, and check in turn whether the routing entries can be used for message forwarding;

[0012] S4, message backtracking: If message forwarding cannot be achieved based on the existing primary and backup routing entries, message backtracking is performed, and the message is traced back to the upstream node or discarded based on the QoS policy;

[0013] S5. The upstream node reroutes the backtracking message.

[0014] Furthermore, the structure of the routing table in step S1 includes a routing prefix tree, a routing entry set, and an interface status table, specifically:

[0015] Routing prefix tree: stores the index of the routing entry set corresponding to the destination network segment;

[0016] Route entry set: stores multiple routes for a destination network segment, including the primary and backup route entries for the destination network segment; the route entry contains the next-hop node interface address and interface number. The primary route in the route entry set is placed at the front of the set, indicating the highest priority;

[0017] Interface status table: In a satellite network, each satellite has a corresponding outbound interface according to the multiple links configured. Therefore, an interface status table is maintained to save the on / off status of each interface and the corresponding link, represented by UP and DOWN respectively.

[0018] Furthermore, the message forwarding described in step S3 specifically includes the following sub-steps:

[0019] S3.1 For the destination IP address in the parsed message, if the destination IP address is the current node IP address, the current node is the destination node, the message is received successfully, and jumps to S3.3; if the destination IP address is not the current node IP address, the routing table is searched according to the parsed destination IP address, the primary and backup routing entries and the corresponding outbound interface status of the routing entries are obtained, and the following steps are performed in sequence:

[0020] S3.1.1 Check the status of the outbound interface corresponding to the currently selected routing entry:

[0021] If the outbound interface status is DOWN, the corresponding forwarding link fails, jump to step S3.2;

[0022] If the outbound interface status is UP, determine whether the current outbound interface number is the same as the inbound interface number. If they are different, jump to S3.1.2 to forward the message. If they are the same, jump to S3.2.

[0023] S3.1.2 sets the outbound interface and destination MAC address in sequence according to the selected valid routing entry, completes message forwarding, and jumps to S3.3;

[0024] S3.2 Select the next routing entry: If the next routing entry exists, jump to S3.1.1; otherwise jump to S4 and enter the message backtracking stage;

[0025] S3.3 Message forwarding is completed.

[0026] Furthermore, the message backtracking described in step S4 specifically includes the following steps:

[0027] Check the priority field in the IP packet header and decide whether to backtrack or discard the packet based on the QoS policy according to the different priorities and link load rates:

[0028] If the message is backtracked: the source MAC address is used as the destination MAC address, the inbound interface is used as the outbound interface, and the message is forwarded back to the upstream node, and the message forwarding is completed;

[0029] If the message is discarded: discard the current message directly.

[0030] Furthermore, the priority field of the IP packet header is implemented using the TOS (Type of Service) field in the IPv4 packet header or the Traffic Class (traffic type) field in the IPv6 packet header, and is used to define the traffic type and priority of the packet. Specifically, DSCP (Differentiated Services Code Point DSCP) is used for marking and identification. The commonly used DSCP values ​​are as follows:

[0031] BE (Default, Best Effort), the value is 000 000, indicating a normal message;

[0032] AF (Assured Forwarding), represented by AFx1y, where x1 is represented by the first three bits and takes values ​​from 1 to 4, indicating different traffic types. y is represented by the fourth and fifth bits and takes values ​​from 1 to 3, indicating different priorities.

[0033] EF (Expedited Forwarding), with a value of 101 110, is used to carry voice traffic. Voice traffic requires low latency, low jitter, and low packet loss rate, and is the most important message after protocol messages.

[0034] CS (Class Selector) is represented by CSx2. x2 is represented by the first three bits and has a value of 1 to 7. CS6 and CS7 are used for protocol messages and are the highest priority messages in the hardware queue.

[0035] Furthermore, based on the QoS policy, the message is traced back to the upstream node or discarded. Specifically, the traceback policy is set based on different service types and the current link load status. It is defined as a function to determine whether to trace back or discard the packet: the function is set to determine the message processing method when the DSCP is a certain service type and according to the link load rate threshold:

[0036] If the DSCP is BE and the link load rate is ≥α, the packet is discarded directly;

[0037] If the DSCP value is AF and the link load ratio is ≥β, the packet is directly discarded. If the link load ratio is <β, the packet is traced back.

[0038] If the DSCP value is EF and the link load ratio is ≥γ, the packet is directly discarded. If the link load ratio is <γ, the packet is traced back.

[0039] If the DSCP is CS6 / CS7, the packet is directly traced back for any link load rate;

[0040] The link load rate thresholds α, β, and γ should satisfy the condition α<β<γ.

[0041] Furthermore, the QoS policy-based message tracing back to the upstream node or discarding it will incur additional overhead and reduce forwarding performance. A simpler message processing method can be used: if the DSCP is BE, all flows are discarded; if the DSCP is AF, flows are randomly discarded and the remaining messages are traced back; if the DSCP is EF / CS6 / CS7, all flows are traced back.

[0042] Furthermore, the message is traced back to the upstream node. If the message encounters a link failure again during the subsequent forwarding process, it can still be traced back again. If the message encounters a large range of continuous node or link failures, in extreme cases, the message will eventually cause a loop after multiple backtracking and forwarding. Since the TTL (Time To Live) field in the IP message header can limit the number of times the message is forwarded, the message that causes the loop can also be discarded based on this.

[0043] Beneficial Effects

[0044] 1. The present invention proposes an innovative fast rerouting method with a backtracking mechanism and a QoS policy, which can effectively deal with the problem of multi-link failures that frequently occur in a satellite network environment, and can ensure successful transmission of traffic when multi-link failures occur in multiple locations in the network. The present invention is applicable to IPv4 or IPv6. Considering that the communication resources of the satellite network are limited, and the backtracking operation will occupy additional network resources and generate delays, the present invention uses the TOS field of IPv4 or the Traffic Class field of IPv6 to distinguish the service type, and performs dynamic routing in combination with the network load status.

[0045] 2. This mechanism allows the system to quickly return to the previous valid node when both the primary and backup communication paths fail. This method significantly improves the system's adaptability to network interruptions and ensures that data transmission still maintains a high transmission success rate in the event of multiple link failures.

[0046] 3. The present invention adopts an ingress sensing method and records the message ingress interface and source MAC address at the same time. The ingress interface and source MAC address are used as the forwarding egress interface and destination MAC address during the backtracking process, making message backtracking simple and efficient. The recorded ingress interface is also used to prevent routing loops. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 is the routing table entry structure;

[0048] Figure 2 It is the inclined orbit constellation network model;

[0049] Figure 3 for Figure 2 A magnified view of the M part;

[0050] Figure 4 It is a mesh network topology and link failure scenario;

[0051] Figure 5 The routing table entry on node A to the destination network segment D;

[0052] Figure 6 It is the routing table entry on node S to the destination network segment D;

[0053] Figure 7 Forwarding flow chart for nodes. DETAILED DESCRIPTION

[0054] The following further describes how the present invention can be applied in conjunction with a low-orbit inclined orbit satellite network scenario. Figure 2 The inclined orbit constellation network model is shown. Figure 3 for Figure 2 The enlarged view of the M part in the figure shows that according to the connection relationship of the intersatellite links, the network topology can be abstracted as follows: Figure 4 The mesh topology shown in Figure 1. In this structure, there may be multiple load-sharing paths between each pair of source and destination nodes, which provides an ideal basis for building backup paths.

[0055] The present embodiment provides a fast rerouting method with a backtracking mechanism and a QoS policy, the method comprising the following specific steps:

[0056] S1 Preprocessing: Each node in the satellite network is configured with a primary and backup routing table. The routing table structure includes the following parts:

[0057] Routing prefix tree: stores the index of the routing entry set corresponding to the destination network segment;

[0058] Route entry set: stores multiple routes for a destination network segment, including the primary and backup route entries for the destination network segment; the route entry contains the next-hop node interface address and interface number. The primary route in the route entry set is placed at the front of the set, indicating the highest priority;

[0059] Interface status table: In a satellite network, each satellite has a corresponding outbound interface according to the multiple links configured. Therefore, an interface status table is maintained to save the on / off status of each interface and the corresponding link, represented by UP and DOWN respectively.

[0060] The specific routing entry set is as follows Figure 1As shown: Find the corresponding set of routing entries from the destination network segment. Each routing entry should contain the next hop node interface IP address and the outbound interface number, and find the interface status table to obtain the outbound interface status. The preprocessing step only needs to be performed once and does not need to be repeated later.

[0061] After receiving the message, the S2 node will perform the following processing steps:

[0062] S2.1 Node ingress perception: When receiving a message, record the ingress interface number and source MAC address of the message;

[0063] S2.2 parses the destination IP address in the message. If the destination node IP address is the current node IP address, the current node is the destination node, the message is received successfully, and jumps to S5;

[0064] S3 searches the routing table according to the parsed destination IP address, obtains the primary and backup routing entries and the status of the outbound interfaces corresponding to the routing entries, and performs the following steps in sequence:

[0065] S3.1 Check the status of the outbound interface corresponding to the currently selected routing entry:

[0066] If the outbound interface status in S3.1.1 is DOWN, the corresponding forwarding link fails and the process jumps to step S3.3;

[0067] S3.1.2 If the outbound interface status is UP, determine whether the current outbound interface number is the same as the inbound interface number. If they are different, jump to S3.2 to forward the message. If they are the same, jump to S3.3.

[0068] S3.2 Message forwarding: According to the selected valid routing entry, the outbound interface and destination MAC address are set in sequence, and the message forwarding is completed, and the process jumps to S5;

[0069] S3.3 Select the next routing entry: If the next routing entry exists, jump to S3.1; otherwise jump to S4 and enter the message backtracking stage;

[0070] S4 message backtracking: Check the priority field of the IP message header, and decide whether to backtrack or discard the message based on the QoS policy according to different priorities and link load rates:

[0071] S4.1 If the message is backtracked: use the source MAC address as the destination MAC address, use the inbound interface as the outbound interface, forward the message back to the upstream node, and jump to S5;

[0072] S4.2 If the message is discarded: directly discard the current message;

[0073] S5 message forwarding is completed.

[0074] In step S3.1.2, it is determined that the current outgoing interface number is the same as the incoming interface number, and the routing entry is skipped to avoid routing loops.

[0075] In step S3.3, if the next routing entry selected does not exist, it means that both the primary and backup routing entries cannot complete message forwarding, and backtracking processing is required for the current message.

[0076] The priority field of the IP packet header mentioned in step S4 is implemented using the TOS (Type of Service) field in the IPv4 packet header or the Traffic Class field in the IPv6 packet header, and is used to define the traffic type and priority of the packet.

[0077] This method is designed to meet diverse business needs by segmenting network traffic into different categories and marking and identifying them using DSCP (Differentiated Services Code Point) values. Commonly used DSCP values ​​are as follows:

[0078] BE (Default, Best Effort), the value is 000 000, indicating a normal message;

[0079] AF (Assured Forwarding) is expressed as AFx1y. x1 is represented by the first three bits and takes values ​​from 1 to 4, indicating different traffic types. y is represented by the fourth and fifth bits and takes values ​​from 1 to 3, indicating different priorities.

[0080] EF (Expedited Forwarding), with a value of 101 110, is used to carry voice traffic. Voice traffic requires low latency, low jitter, and low packet loss rate, and is the most important message after protocol messages.

[0081] CS (Class Selector) can be expressed as CSx2, where x2 is represented by the first three bits and takes values ​​from 1 to 7. CS6 and CS7 are usually used for protocol messages. If these messages cannot be received, it will cause a protocol interruption. Therefore, they are the highest priority messages in the hardware queues of most manufacturers.

[0082] When both the primary and backup paths fail, a forwarding strategy is set for the node. A backtracking strategy is set based on different service types and the current link load status. The backtracking strategy is defined as a function to determine whether to backtrack or drop packets:

[0083] A typical setting method is: when DSCP is a certain service type, it can be set to decide whether to discard the message based on the link load rate threshold:

[0084] If the DSCP is BE and the link load rate is ≥α, the packet is discarded directly;

[0085] If the DSCP value is AF and the link load ratio is ≥β, the packet is directly discarded. If the link load ratio is <β, the packet is traced back.

[0086] If the DSCP value is EF and the link load ratio is ≥γ, the packet is directly discarded. If the link load ratio is <γ, the packet is traced back.

[0087] If the DSCP is CS6 / CS7, message traceback is performed directly for any link load rate.

[0088] The load thresholds α, β, and γ should satisfy the condition α<β<γ, which reflects the priority order of service types, where CS6 / CS7 has the highest priority, followed by EF, then AF, and BE has the lowest priority;

[0089] Considering that obtaining the link load status will incur additional overhead and reduce forwarding performance, a simpler backtracking strategy can be adopted: discard all BE flows; randomly discard AF flows and backtrack the remaining messages; and backtrack all EF / CS6 / CS7 flows.

[0090] If the message encounters a link failure again during the subsequent forwarding process, it can still be traced back again. When a large range of continuous node link failures occur, the message undergoes multiple backtracking and forwarding, and in extreme cases, a loop may occur. In this case, the TTL field in the IP packet header can also be used to limit the number of times the message is forwarded, and the message that may cause a loop will be discarded.

[0091] This ensures that even when multiple nodes and multiple links fail, the system can still ensure high reliability of transmission through repeated backtracking.

[0092] exist Figure 4 In the network topology, the destination node is marked as D. Node A Figure 5 The routing table shown forwards the message to node S, and the three links around S fail.

[0093] The DSCP value of the message can be set as listed in Table 1. Assume that the messages arriving at node S are two types: enterprise internal common data flow services and video conferencing service messages, and their DSCP values ​​are AF11 and EF respectively:

[0094] Table 1 DSCP value settings for packets

[0095]

[0096] The QoS policy of the message traceback mechanism is listed in Table 2:

[0097] Table 2 QoS policy of message traceback mechanism

[0098]

[0099] Figure 4 The IP addresses of the main nodes A, S, and D in the network are assigned as follows:

[0100] Node A:

[0101] IP address of interface 1 (connected to node 6): 10.1.0.2 / 24;

[0102] IP address of interface 2 (connected to node 1): 10.1.1.1 / 24;

[0103] IP address of interface 3 (connected to node S): 10.1.2.1 / 24;

[0104] IP address of interface 4 (connected to node 5): 10.1.3.1 / 24;

[0105] Node S:

[0106] IP address of interface 1 (connected to node A): 10.1.2.2 / 24;

[0107] IP address of interface 2 (connected to node 2): 10.1.3.1 / 24;

[0108] IP address of interface 3 (connected to node 3): 10.1.4.1 / 24;

[0109] IP address of interface 4 (connected to node 4): 10.1.5.1 / 24;

[0110] The IP address of node D is 10.1.6.2 / 24 and the network segment is 10.1.6.0 / 24.

[0111] First create a routing table entry for the network, such as Figure 5 and Figure 6 The following are the destination network segment routing entries on nodes A and S that are sent to D.

[0112] When node S receives the message, it executes Figure 7 Shown behavior:

[0113] Node S first records the incoming interface (interface 1) and the source MAC address;

[0114] Find the routing table entry set corresponding to the destination network segment, such as Figure 6 As shown;

[0115] Check the outbound interface status from top to bottom in the routing table set. If the status is DOWN or the inbound interface number is equal to the outbound interface number, select the next routing table entry.

[0116] For node S, if Figure 6 It can be seen that the first three interface states in the routing table are all DOWN. The last routing table entry is not considered because the outgoing interface (interface 1) is equal to the incoming interface (interface 1), so it will enter the backtracking state;

[0117] Get the DSCP value in the packet header, check the link load rate of the traceback path, and decide whether to trace the packet based on Table 1 and Table 2:

[0118] Assuming that the link load rate between nodes A and S is 70% at this time, referring to Table 2, it can be seen that the video conference service message can be traced back, while the ordinary data flow service message within the enterprise will be discarded.

[0119] For the traceback message, the source MAC address is set to the destination MAC address, the inbound interface is set to the outbound interface, and the message is forwarded to the upstream node A.

[0120] When node A receives the backtrace message, it will also execute the same processing flow as S:

[0121] Record the inbound interface (interface 3) and source MAC address;

[0122] Find the routing table entry set corresponding to the destination network segment, such as Figure 5 As shown;

[0123] Check the outbound interface status from top to bottom in the routing table set. If the status is DOWN or the inbound interface number is equal to the outbound interface number, select the next routing table entry.

[0124] For node A, if Figure 5 As shown, the first entry in the routing table is taken out and checked: Since the outgoing interface (interface 3) of the routing table entry is equal to the incoming interface (interface 3), the routing table entry is ignored and the next one is selected;

[0125] Since the second routing entry meets the forwarding conditions, the message is forwarded directly.

[0126] During the forwarding process of node A, it uses the inbound interface to prevent the message from returning to node S again, effectively preventing the routing loop problem. The specific performance of node A's re-routing process is as follows Figure 4 As shown in the figure, there are two equal-cost paths from A to D, where P1 is the main path and P2 is the load-balancing path. Since P1 is the backtracking path of S, A will choose P2 to implement message forwarding, thereby avoiding link failure on node S.

[0127] In the subsequent message forwarding, if the link failure occurs again, the above steps will still be performed; when a large range of continuous node link failures occur, the message undergoes multiple backtracking and forwarding, and a loop may occur in extreme cases. In this case, the TTL field in the IP packet header can also be used to limit the number of times the message is forwarded, and the message that may cause a loop will be discarded.

[0128] In general, the fast rerouting method with backtracking mechanism and QoS strategy proposed in the present invention can cope with multi-node and multi-link failure phenomena and improve the reliability of the network.

Claims

1. A fast rerouting method with a backtracking mechanism and QoS strategy, characterized in that: The method comprises the following steps: S1. Preprocessing: Each node in the satellite network is configured with a primary and backup routing table. The corresponding routing entry set is found by the destination network segment. Each routing entry should contain the next hop node interface IP address and the outbound interface number, and the interface status table is searched to obtain the outbound interface status. S2, message reception: When a node receives a message, it records the incoming interface number and source MAC address of the message, and parses the destination IP address in the message; S3, message forwarding: After receiving the message, the node will search the routing table according to the parsed destination IP address, obtain the primary and backup routing entries and the status of the corresponding outbound interfaces of the routing entries, and check in turn whether the routing entries can be used for message forwarding; S4, message backtracking: If message forwarding cannot be achieved based on the existing primary and backup routing entries, message backtracking is performed, and the message is traced back to the upstream node or discarded based on the QoS policy; the message backtracking specifically includes the following steps: Check the priority field in the IP packet header and decide whether to backtrack or discard the packet based on the QoS policy according to the different priorities and link load rates: If the message is backtracked: the source MAC address is used as the destination MAC address, the inbound interface is used as the outbound interface, and the message is forwarded back to the upstream node, and the message forwarding ends; If the message is discarded: directly discard the current message; The priority field of the IP packet header is implemented using the TOS field in the IPv4 data packet header or the Traffic Class field in the IPv6 data packet header, and is used to define the traffic type and priority of the packet. Specifically, DSCP is used for marking and identification. The DSCP value is as follows: BE, value is 000 000, representing a normal message; AF, expressed as AFx1y, where x1 is represented by the first three bits and takes values ​​from 1 to 4, indicating different traffic types. y is represented by the fourth and fifth bits and takes values ​​from 1 to 3, indicating different priorities. EF, with a value of 101 110, is used to carry voice traffic. Voice traffic requires low latency, low jitter, and low packet loss rate, and is the most important message after protocol messages. CS, represented by CSx2, where x2 is represented by the first three bits and has a value of 1 to 7. CS6 and CS7 are used for protocol messages and are the highest priority messages in the hardware queue; S5. The upstream node reroutes the backtracking message.

2. The fast rerouting method with backtracking mechanism and QoS strategy according to claim 1, characterized in that: The structure of the routing table in step S1 includes a routing prefix tree, a routing entry set, and an interface status table, specifically: Routing prefix tree: stores the index of the routing entry set corresponding to the destination network segment; Route entry set: stores multiple routes for a destination network segment, including the primary and backup route entries for the destination network segment; The routing entry contains the next-hop node interface address and interface number. The main route is placed at the front of the set of routing entries, indicating the highest priority. Interface status table: In a satellite network, each satellite has a corresponding outbound interface according to the multiple links configured. Therefore, an interface status table is maintained to save the on / off status of each interface and the corresponding link, represented by UP and DOWN respectively.

3. The fast rerouting method with backtracking mechanism and QoS strategy according to claim 1, characterized in that: The message forwarding described in step S3 specifically includes the following sub-steps: S3.1 For the destination IP address in the parsed message, if the destination IP address is the current node IP address, the current node is the destination node, the message is received successfully, and jumps to S3.3; If the destination IP address is not the current node IP address, the routing table is searched according to the parsed destination IP address to obtain the primary and backup routing entries and the status of the outbound interfaces corresponding to the routing entries, and the following steps are performed in sequence: S3.1.1 Check the status of the outbound interface corresponding to the currently selected routing entry: If the outbound interface status is DOWN, the corresponding forwarding link fails, jump to step S3.2; If the outbound interface status is UP, determine whether the current outbound interface number is the same as the inbound interface number. If they are different, jump to S3.1.2 to forward the message. If they are the same, jump to S3.

2. S3.1.2 According to the selected valid routing entry, the outbound interface and destination MAC address are set in sequence, and the message forwarding is completed, and the process jumps to S3.3; S3.2 Select the next routing entry: If the next routing entry exists, jump to S3.1.1; otherwise jump to S4 and enter the message backtracking stage; S3.3 Message forwarding is completed.

4. The fast rerouting method with backtracking mechanism and QoS strategy according to claim 1, characterized in that: Based on the QoS policy, the message is traced back to the upstream node or discarded. Specifically, the traceback policy is set based on different service types and the current link load status. It is defined as a function to determine whether to trace back or discard the packet. The function is set to determine the message processing method when the DSCP is a certain service type and based on the link load rate threshold: If the DSCP is BE and the link load rate is ≥α, the packet is discarded directly; If the DSCP value is AF and the link load ratio is ≥β, the packet is directly discarded. If the link load ratio is <β, the packet is traced back. If the DSCP value is EF and the link load ratio is ≥γ, the packet is directly discarded. If the link load ratio is <γ, the packet is traced back. If the DSCP is CS6 / CS7, the packet is directly traced back for any link load rate; The link load rate thresholds α, β, and γ should satisfy the condition α<β<γ.

5. The fast rerouting method with backtracking mechanism and QoS strategy according to claim 1, characterized in that: The QoS policy-based message is traced back to the upstream node or discarded, and the message processing method adopted is: if the DSCP is BE flow, all flows are discarded; if the DSCP is AF flow, the flow is randomly discarded and the remaining messages are traced back; if the DSCP is EF / CS6 / CS7 flow, all flows are traced back.

6. The fast rerouting method with backtracking mechanism and QoS strategy according to claim 1, characterized in that: The message is traced back to the upstream node and the TTL field in the IP message header is used to limit the number of forwarding times.

Images