An ITUT G.8032 optimization method based on traffic adaptation

Abstract

The present invention belongs to the field of computer application technology, and provides an ITUTG.8032 optimization method based on traffic adaptation. According to the existing ERPS protocol, when facing bandwidth changes between nodes in a loop, RPL selection is only performed during initial planning, and the redundant links cannot be fully utilized to share traffic load. The present invention proposes an ITUTG.8032 optimization method based on traffic adaptation based on the feature that ERPS supports multi-instance configuration. While fully utilizing the multi-instance configuration feature of ERPS, this method creates a corresponding instance for each node and manages the data flow of each node in detail. When a link configured with multiple ERPS instances is affected by bandwidth reduction, it will be calculated based on the communication quality pre-configuration to determine whether to select the RPL of a certain instance on the link as the link, so as to avoid the result of communication quality deterioration due to the reduction of the link bandwidth, thereby effectively reducing the network delay of each node in the topology and improving communication quality.

Description

Technical Field

[0001] The invention belongs to the technical field of computer application, and provides an ITUT G.8032 optimization method based on flow adaptation, and an improved method for applying a multi-instance configuration method based on the flow adaptation design principle of the ITUT G.8032 Ethernet ring network protection protocol. Background Art

[0002] The ITUT G.8032 protocol (ERPS protocol - Ethernet Ring Protection Switching), as an Ethernet ring network protection protocol, is widely used for its fast convergence speed and ability to meet the carrier-class switching time requirement of 50ms. To ensure that the entire network continues to operate normally when a single link fails, link redundancy is typically implemented. To prevent topological loops, the ERPS protocol blocks a single link. This blocked link is called the RPL (Ring Protection Link). In this protection state, once a link failure is detected, the faulty link is blocked and the previously blocked RPL is released to implement fault protection switching.

[0003] In the ERPS protocol, multiple nodes are connected end to end to form a physical loop. A physical loop can have one or more instances, each of which can be configured with its own control VLAN and RPL. Each ERPS instance uses a different control VLAN to transmit its corresponding RAPS message. Although the original intention of ERPS to support multi-instance configuration is to fully utilize redundant links, the RPL in each instance configuration is only manually configured at the beginning of the configuration, and may not fully achieve traffic load sharing as the effective bandwidth of the link changes. Therefore, we propose an ERPS protocol optimization configuration method based on traffic adaptation, which can dynamically adjust the position of the RPL according to the traffic changes of each node to achieve better traffic load sharing, effectively reduce the network delay of each node in the topology, and improve communication quality. Summary of the Invention

[0004] According to the existing ERPS protocol, when facing bandwidth changes between nodes in the loop, RPL selection is only performed during initial planning, which makes it impossible to fully utilize redundant links for traffic load sharing. In order to solve the problems existing in the above-mentioned prior art means, the present invention proposes a traffic-adaptive ITUT G.8032 optimization method based on the feature that ERPS supports multi-instance configuration. While fully utilizing the multi-instance configuration feature of ERPS, this method dynamically selects RPL based on the effective bandwidth between each node by regularly calculating the throughput between each node, thereby being able to dynamically adjust the position of RPL according to the traffic changes of each node to achieve better traffic load sharing.

[0005] The technical solution of the present invention is:

[0006] An ITUT G.8032 optimization method based on traffic adaptation includes the following steps:

[0007] (1) First, in a physical loop with N nodes, create a corresponding ERPS instance for each node in the loop except the traffic convergence point, and configure the maximum acceptable attenuation threshold C according to the communication quality. max and the minimum total flow threshold T at the flow convergence point min ;

[0008] (2) Select S in each link n -B n Maximum link L n , in L n Select instances from the traffic flow, and adjust the RPL of each instance according to the total traffic value of each instance at the traffic convergence point until the total traffic value at the traffic convergence point is greater than the initial set minimum total traffic threshold T min Among them, L n represents the link between adjacent nodes n and n+1, B n represents the effective bandwidth of the link between adjacent nodes n and n+1. The set of all instances of the link between adjacent nodes n and n+1 is I, i∈I, T i (L n ) represents the traffic from the link between adjacent nodes n and n+1 in the i-th instance, then the total traffic between the links n and n+1 is S n =∑ i∈I T i (L n );

[0009] (3) During the operation of the loop, a link L appears n When signal attenuation occurs due to channel quality changes, determine whether the difference between the traffic demand and the actual bandwidth of the link where signal attenuation occurs is greater than C max If yes, make adaptive adjustments.

[0010] Among them, the specific method of step (1) is:

[0011] (101) The N nodes in the physical loop are numbered clockwise starting from 1, starting from the right neighboring node of the traffic convergence point. Each node n i Corresponding to an instance i, a different control VLAN is configured for each instance;

[0012] (102) In the initial state, each instance selects the link with the smallest effective bandwidth as RPL for blocking. Assume that the two ends of the selected RPL are n r and n r+1 and denote the effective bandwidth of the selected link as B rpl .

[0013] Among them, the specific method of step (2) is:

[0014] (201) Select S in each link n -B n Maximum link L n , in L n Select an instance, traverse instance i in turn, and calculate the node n corresponding to instance i i and n i+1 The total traffic value at the traffic convergence point after the link between them is selected as RPL;

[0015] (202) If the total flow value is the largest when a certain instance j is selected, then the RPL position corresponding to instance j is changed. Assuming that the initial n r If r>j, then go counterclockwise to node n. j Select any link between the two nodes as RPL to change the traffic direction;

[0016] (203) If there is no instance j that makes the total flow after the change greater than the original total flow, then terminate the RPL adjustment and directly execute step (3); if there is, execute step (204);

[0017] (204) Traverse the traffic of the links between each node and calculate the total traffic S between the links of node n and n+1 after the RPL change in instance j n =∑ i∈I T i (L n );

[0018] (205) If the total flow of the traffic aggregation node is not greater than the initial minimum total flow threshold T min And there is a link that satisfies S n >B n, then go to step (201) until the total flow value at the flow convergence point is greater than the minimum total flow threshold T min , the RPL of each instance will no longer be adjusted, and the maximum number of adjustments is no more than 2 N-1 Otherwise, go directly to step (3).

[0019] Among them, the specific method of step (3) is:

[0020] (301) When the link between nodes n and n+1 experiences signal attenuation, determine whether the difference between the traffic demand and the actual bandwidth of the link experiencing signal attenuation is greater than C max , when S n -B n >C max When , enter step (302) to start adaptive adjustment calculation;

[0021] (302) In link L n Traverse and select instance k from the involved instances, so that |S n -B n -T k (L n )| is the smallest. If the RPL position of instance k is changed, the link L involved in the new traffic direction will be changed after the traffic direction is changed. m Is there a flow demand greater than B? m Links;

[0022] (303) If it does not exist, release the original RPLowner port of instance k, select the new RPL as the link with signal attenuation, and send RPL blocking RAPS message; if it exists, compare S n -B n and S m -B m If the size of S n -B n If S is larger, the original RPL owner port of instance k is still released, the new RPL is selected as the link with signal attenuation, and the RPL blocking RAPS message is sent; if S m -B m If it is larger, adaptive adjustment is abandoned.

[0023] The beneficial effects of the present invention are as follows: By leveraging the configurable multi-instance feature of the ERPS ring, this method creates a corresponding instance for each node, meticulously managing the data flow of each node. Furthermore, the proposed routing solution can reduce the load on the ring edge and achieve traffic load sharing. When a link configured with multiple ERPS instances is affected by bandwidth reduction, a calculation is performed based on the maximum acceptable attenuation threshold configured for communication quality and the minimum total traffic threshold at the traffic convergence point. Based on the calculation result, it is determined whether to select the RPL of a specific instance on that link as that link. This avoids the result of communication quality degradation caused by the bandwidth reduction of that link, effectively reducing network latency for each node in the topology and improving communication quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 is a flow chart of the method of the present invention;

[0025] Figure 2 It is a graphic representation of the RPL selection effect of multiple instances of the present invention. DETAILED DESCRIPTION

[0026] In order to make the purpose, technical solution and advantages of the present invention more clear, the following is a clear and complete description of the technical solution in the implementation of the present invention in conjunction with the accompanying drawings. Figure 1 As shown:

[0027] First, create a corresponding ERPS instance for each node in the loop except the traffic aggregation point, and complete the initial configuration for each instance.

[0028] (1) First, in a physical loop with N nodes, create a corresponding ERPS instance for each node in the loop except the traffic convergence point; Figure 2 As shown in the figure, in a ring with five nodes, we create four ERPS instances for each of these five nodes, excluding the traffic aggregation point. The instance corresponding to node N3 is represented by a short dashed line, and the instance corresponding to node N4 is represented by a long dashed line. Configure the maximum acceptable attenuation threshold C based on the communication quality. max and the minimum total flow threshold T at the flow convergence point min The specific method of step (1) is:

[0029] (101) The N nodes in the physical loop are numbered clockwise starting from 1, starting from the right neighboring node of the traffic convergence point. Each node n i Corresponding to an instance i, a different control VLAN is configured for each instance;

[0030] (102) In the initial state, each instance selects the link with the smallest effective bandwidth as RPL for blocking. Let the two ends of the selected RPL be nr and n r+1 and denote the effective bandwidth of the selected link as B rpl .

[0031] Second, traffic simulation is performed based on the effective bandwidth between nodes and the service traffic planned for each node, and a better RPL is calculated for each instance to fully utilize the link.

[0032] (2) Select S in each link n -B n Maximum link L n , in L n Select instances from the traffic flow, and adjust the RPL of each instance according to the total traffic value of each instance at the traffic convergence point until the total traffic value at the traffic convergence point is greater than the initial set minimum total traffic threshold T min Among them, L n represents the link between adjacent nodes n and n+1, B n represents the effective bandwidth of the link between n and n+1. The set of all instances of the link between adjacent nodes n and n+1 is I, i∈I, T i (L n ) represents the traffic from the link between adjacent nodes n and n+1 in the i-th instance, then S n =∑ i∈I T i (L n ) is the total traffic between links n and n+1. The specific method of step (2) is:

[0033] (201) Select S in each link n -B n Maximum link L n , in L n Select an instance, traverse instance i in turn, and calculate the node n corresponding to instance i i and n i+1 The total traffic value at the traffic convergence point after the link between them is selected as RPL;

[0034] (202) If the total flow value is the largest when instance j is selected, then the RPL position corresponding to instance j is changed. Assuming that the initial n r If r>j, then go counterclockwise to node n. j Select any link between the two nodes as RPL to change the traffic direction;

[0035] (203) If there is no instance j that makes the total flow after the change greater than the original total flow, then terminate the RPL adjustment and directly execute step (3); if there is, execute step (204);

[0036] (204) Traverse the traffic of the links between each node and calculate the total traffic S between the links of node n and n+1 after the RPL change in instance j n =∑ i∈I T i (L n );

[0037] (205) If the total flow of the traffic aggregation node is not greater than the initial minimum total flow threshold T min And there is a link that satisfies S n >B n , then go to step (201) until the total flow value at the flow convergence point is greater than the minimum total flow threshold T min , the RPL of each instance will no longer be adjusted, and the maximum number of adjustments is no more than 2 N-1 Otherwise, go directly to step (3).

[0038] Third, when a link L appears during the operation of the loop n Since the signal attenuation is caused by the change in channel quality, it is determined whether the link RPL needs to be adaptively adjusted according to the change in channel quality.

[0039] (3) After completing the above configuration, the loop can work normally. n Since the signal attenuation is caused by the change in channel quality, it is determined whether to perform adaptive adjustment of the link RPL according to the change in channel quality. The specific method of step (3) is as follows:

[0040] (301) When the link between nodes n and n+1 experiences signal attenuation, determine whether the difference between the traffic demand and the actual bandwidth of the link experiencing signal attenuation is greater than C max , when S n -B n >C max When , enter step (302) to start adaptive adjustment calculation;

[0041] (302) In link L n Traverse and select instance k from the involved instances, so that |S n -B n -T k (L n )| is the smallest. If the RPL position of instance k is changed, the link L involved in the new traffic direction will be changed after the traffic direction is changed. m Is there a flow demand greater than B? m Links;

[0042] (303) If it does not exist, release the original RPLowner port of instance k, select the new RPL as the link with signal attenuation, and send RPL blocking RAPS message; if it exists, compare S n -B n and S m -B m If the size of S n -B n If S is larger, the original RPL owner port of instance k is still released, the new RPL is selected as the link with signal attenuation, and the RPL blocking RAPS message is sent; if S m -B m If it is larger, adaptive adjustment is abandoned.

[0043] like Figure 2 As shown, for example, the link between N4 and N5 has signal attenuation. After the judgment in step (3), the instance corresponding to N3 is selected for RPL adjustment, and the new RPL is selected as the link with signal attenuation. The data flow state is adjusted from the left figure to the right figure.

Claims

1. An ITUT G.8032 optimization method based on traffic adaptation, characterized in that: Here are the steps: (1) First, in a physical loop with N nodes, a corresponding ERPS instance is created for each node in the loop except the traffic convergence point. In the initial state, each instance selects the link with the smallest effective bandwidth as RPL for blocking, and configures the maximum acceptable attenuation threshold C according to the communication quality. max and the minimum total flow threshold T at the flow convergence point min ; (2) Select S in each link n -B n Maximum link L n , in L n Select instances from the traffic flow, and adjust the RPL of each instance according to the total traffic value of each instance at the traffic convergence point until the total traffic value at the traffic convergence point is greater than the initial minimum total traffic threshold T min Among them, L n represents the link between adjacent nodes n and n+1, B n represents the effective bandwidth of the link between adjacent nodes n and n+1. The set of all instances of the link between adjacent nodes n and n+1 is I, i∈I, T i (L n ) represents the traffic from the link between adjacent nodes n and n+1 in the i-th instance, then the total traffic between the links n and n+1 is S n =∑ i∈ I T i (L n ); (3) During the operation of the loop, a link L appears n When signal attenuation occurs due to channel quality changes, determine whether the difference between the traffic demand and the actual bandwidth of the link where signal attenuation occurs is greater than C max , if so, make adaptive adjustments; Among them, the specific method of step (3) is: (301) When the link between nodes n and n+1 experiences signal attenuation, determine whether the difference between the traffic demand and the actual bandwidth of the link experiencing signal attenuation is greater than C max , when S n -B n >C max When , enter step (302) to start adaptive adjustment calculation; (302) In link L n Traverse and select instance k from the involved instances, so that |S n -B n -T k (L n )| is the smallest. If the RPL position of instance k is changed and the traffic direction is changed, the link L involved in the new traffic direction m Is there a flow demand greater than B? m Links; (303) If it does not exist, release the original RPLowner port of instance k, select the new RPL as the link with signal attenuation, and send RPL blocking RAPS message; if it exists, compare S n -B n and S m -B m If the size of S n -B n If S is larger, the original RPL owner port of instance k is still released, the new RPL is selected as the link with signal attenuation, and the RPL blocking RAPS message is sent; if S m -B m If it is larger, adaptive adjustment is abandoned.

2. The ITUT G.8032 optimization method based on traffic adaptation according to claim 1, characterized in that: The specific method of step (1) is: (101) The N nodes in the physical loop are numbered clockwise starting from 1, starting from the right neighboring node of the traffic convergence point. Each node n i Corresponding to an instance i, a different control VLAN is configured for each instance; (102) In the initial state, each instance selects the link with the smallest effective bandwidth as RPL for blocking. Assume that the two ends of the selected RPL are n r and n r+1 and denote the effective bandwidth of the selected link as B rpl .

3. The ITUT G.8032 optimization method based on traffic adaptation according to claim 1, characterized in that: The specific method of step (2) is: (201) Select S in each link n -B n Maximum link L n , in L n Select an instance, traverse instance i in turn, and calculate the node n corresponding to instance i i and n i+1 The total traffic value at the traffic convergence point after the link between them is selected as RPL; (202) If the total flow value is the largest when a certain instance j is selected, then the RPL position corresponding to instance j is changed. Assuming that the initial n r If r>j, then go counterclockwise to node n. j Select any link between the two nodes as RPL to change the traffic direction; (203) If there is no instance j that makes the total flow after the change greater than the original total flow, then terminate the RPL adjustment and directly execute step (3); if there is, execute step (204); (204) Traverse the traffic of the links between each node and calculate the total traffic S between the links of node n and n+1 after the RPL change in instance j n =∑ i∈I T i (L n ); (205) If the total flow of the traffic aggregation node is not greater than the initial minimum total flow threshold T min And there is a link that satisfies S n >B n , then go to step (201) until the total flow value at the flow convergence point is greater than the minimum total flow threshold T min , the RPL of each instance will no longer be adjusted, and the maximum number of adjustments is no more than 2 N-1 Otherwise, go directly to step (3).

Images