KR20020053393A - An Efficient Restoration Mechanism Using Bandwidth Sharing Method In MPLS - Google Patents

Abstract

PURPOSE: A method for efficiently restoring a path by bandwidth sharing in an MPLS(Multiprotocol Label Switching) network is provided to improve the efficiency of a network and maintain the high reliability of the network by enabling restoration paths to share bandwidth each other providing working paths with more network resources. CONSTITUTION: Network availability information such as each node on a path to be set, links between each node, bandwidth between each link, and the maximum bandwidth of restoration paths between each link, are grasped(301). The nodes having a bandwidth larger than the requested bandwidth are selected in the links(302). The shortest path between the selected nodes is calculated and the calculated path is set as a working path(303-306). If the working path is set, nodes and links except links having the sum of bandwidth between the links and the maximum bandwidth of the restoration path smaller than the requested bandwidth, are selected(307). The shortest path between the selected nodes is calculated and the calculated path is set as the restoration path(308-310).

Description

Efficient Restoration Mechanism Using Bandwidth Sharing Method In MPLS}

The present invention relates to a network recovery method in a network. More specifically, in establishing a recovery path in an MPLS network, the recovery paths share bandwidth with each other to provide more network resources for a work path. The present invention relates to a recovery path establishment method that can increase network efficiency while maintaining high reliability of services provided.

Recently, the increase of Internet users, the increase of traffic, and the diversification of application services through the Internet require the high speed of the Internet and the guarantee of service quality. To address these challenges, MPLS (Multiprotocol Label Switching) technology, which combines the link layer switching technology and delivers IP packets, is being standardized.

MPLS is a technology for reinforcing the Internet function of the Asynchronous Exchange (ATM) exchange, which is the core of the high-speed information communication network. MPLS integrates and transmits layer 3 and layer 2 so that label assignment is performed independently of the layer 3 protocol. In addition to supporting them, it can be flexibly applied to lower transport networks.

When such MPLS technology is used as a backbone network in the future, a stable and efficient path recovery method must be provided.

Path recovery means setting recovery path to maintain data transmission by bypassing the recovery path when a problem occurs during data transmission or normal data transmission is not possible due to an error of a specific node on a preset LSP (Label Switched Path). This is a description of how to recover.

The basic operation concept of MPLS is shown in FIG. 1. The Ingress Label Switching Router (LSR) in the MPLS domain analyzes the header of the packet from the non-MPLS network to determine the LSP to which the packet will be delivered and attach the label according to the corresponding exit interface. The label of the MPLS is a local identification that has meaning only within one LSR. In the middle LSRs, when a labeled packet comes in, only the label is examined, and packet processing and label swapping are performed according to the label. The egress LSR removes the label from the arriving packet and forwards the packet to its destination. At this time, the label lookup table in each LSR is generated by a label distribution protocol (LDP) which is a signaling protocol of MPLS.

As shown in FIG. 1, the MPLS network is composed of an MPLS edge node 101, an MPLS internal node 102, and a link 103 connecting these nodes. An edge node is a node having a connection with an external network, and an inner node is a node having a connection with an MPLS node.

The MPLS network establishes a label switched path (LSP) from an edge node to another edge node by an external request. For example, when the LSP is set from the F edge node to the C edge node by the proposed method, a work path 104 via J, which is the shortest path, is set, and a recovery path 105 via G and H is set. Is established, and if necessary, another recovery path 106 via L, KI is established. The working path is 2 in length, the length of recovery path 1 is 3, and the length of recovery path 2 is 4. These paths are all disjoint paths with different nodes and different links.

Traffic engineering, which is emerging as a key feature of MPLS, is the process of mapping the corresponding requirements of the traffic pack to the network for the purpose of optimizing the utilization of network resources while using network operations efficiently and reliably. This maximizes the utilization of network links and nodes, and also distributes network traffic across multiple links appropriately, minimizing the impact of one link or node failure or ensuring extra link capacity to bypass traffic in the event of a failure. It is possible.

Constrained-Based Routing can be used as an important tool for this traffic engineering, with explicit route selection considering traffic information and network resources.

Meanwhile, network recovery function can be provided at each layer. Network recovery in the optical layer, which performs recovery in wavelength units, enables fast fault recognition and rapid recovery of a large number of higher layer data flows by rapid detection of a transmission signal. In addition, recovery is possible at the SDH / SONET layer, which provides automatic protection switching.

However, the recovery in the optical layer or SDH / SONET layer can not distinguish the traffic for the logical path of the upper layer, there is a disadvantage that the recovery of the network according to the service can not be supported. In addition, in the current Internet, routing information is transmitted when a failure occurs, and it relies on an IP layer network recovery function that delivers a packet to a new path by changed routing information. This method takes a considerable amount of time compared to the lower layer, and has a disadvantage in that an alternative path for recovery cannot be established because there is no explicit path determining function.

However, network recovery in MPLS can provide several advantages over the existing network recovery providing method. In other words, it can detect the failure of logical channel or node configuration error that cannot be found in optical layer or SDH layer, and the explicit path setting considering bandwidth with Constrained-Based Routing function not provided in IP layer. QoS-based network recovery is possible. In addition, the messages provided by the MPLS signaling protocol enable fast detection. However, the failure detection method of MPLS is difficult to detect a physical link failure, and it is insufficient to provide various network recovery functions.

Network recovery methods for providing network recovery in MPLS include the following methods.

In general, every network consists of a node corresponding to a host or a router, a link connecting the nodes, and a path leading to the link. To recover from such a network, a recovery path or protection path is provided along with a working path. There are two methods of setting up a recovery path: a protection switching method in which the recovery path is prepared before the failure, and a rerouting method of finding a new path after the failure.

Recovery by re-routing establishes new paths by referring to fault information, network routing policy and network components after network failure. The rerouting model is slower than protection switching, but is suitable for optimizing resource usage because routing is performed by the current state of the network.

The protection switching model, on the other hand, uses an explicit routing method to pre-establish the recovery path based on traffic recovery information and policy considerations in the working path. When a failure is detected on the work path, the start node of the recovery path immediately replaces the traffic with the recovery path, enabling fast recovery. Such protection switching methods include 1 + 1 ("one plus one") protection and 1: 1 ("one for one") protection depending on how resources of the recovery path are used.

1 + 1 protection transmits the same traffic on the work path and the recovery path in the path switch LSR (PSL) when the path does not cause a problem, and merges the received data without duplication in the path merge LSR (PML). If a path fails, data transfers on the work path are interrupted and data is passed only to the recovery path. This method can receive normal data without any time delay at the destination even if the path fails, but it is less efficient because the resources are doubled.

1: 1 protection does not send the same traffic to the recovery path unless the recovery path is used due to a failure on the work path. Therefore, even if the recovery path is set in advance, when it is in normal operation, it can be used as another low-priority traffic delivery path, thereby effectively utilizing the resources of the recovery path. This concept is extensible with 1: n and m: n protection.

On the other hand, when looking at the recovery structure to the extent of performing recovery, it is divided into local recovery and global recovery.

Partial recovery aims to find a path that bypasses a failed link or node, and the node that detected the failure performs recovery. There are two types of partial recovery: link recovery, which bypasses the failed link, and node recovery, which bypasses the failed node.

The present invention provides a work path and recovery path calculation method for providing fast path recovery while efficiently using network resources in order to ensure the reliability of the network in establishing the network recovery path in the aforementioned MPLS network. It aims to provide.

1 is a schematic diagram of an MPLS network model to which the present invention is applied;

2 is a functional block diagram of an MPLS system including a path recovery mechanism in accordance with the present invention;

3 is a flowchart illustrating a bandwidth sharing recovery path setting mechanism for 1: 1 path recovery in a path recovery method according to the present invention;

4 is a flowchart illustrating a bandwidth sharing recovery path establishment mechanism for 1: n path recovery in a path recovery method according to the present invention; And

5 is a diagram illustrating a method of calculating a bandwidth value in any link used in the path recovery method according to the present invention.

According to a first aspect of the present invention, there is provided an apparatus, comprising: a routing protocol 202 for storing network availability information and routing information; A route calculation block 201 implemented in a source node and calculating a work path and a recovery path based on network availability information stored in the routing protocol 202 in response to a user request; An LDP for setting a path in response to a path setting request from the path calculation block 201 and detecting a failure of a network path; And a path recovery mechanism through a forwarding engine that receives a work path and a recovery path set from the LDP, and forwards a packet to a normal path of the work path or the recovery path through an instruction of the LDP. Is provided.

In addition, according to a second aspect of the present invention, there is provided a method for establishing a working path and a recovery path from one node to another node in an MPLS network, wherein each node on the path to be established is a link between each node. Determining network availability information such as a bandwidth value between each link and a maximum bandwidth value of a recovery path between each link; Selecting a link having a bandwidth greater than the requested bandwidth among the links; A third step of obtaining a shortest path between the nodes selected in the second step and setting it as a work path, and transitioning to the first step if the job setting fails; A fourth step of selecting a node and a link excluding a link in which the sum of the inter-link bandwidth value and the maximum bandwidth value of the recovery path is less than the requested bandwidth value except for the set work path, when the work path is set in the third step; step; A fifth step of obtaining the shortest path between the nodes obtained in the fourth step and setting it as a recovery path; And a sixth step of transitioning to the fourth step if the recovery path is not set when the recovery path is set in the fifth step and failing to set the recovery path.

On the other hand, in the path recovery method, the seventh step of setting the number of recovery paths to an initial value "0" after the working path is set; An eighth step of increasing the number of recovery paths by "1" after the recovery path is established; And a ninth step if the value increased in the eighth step is smaller than the desired number of recovery paths, and transitions to the fourth step, and ends if it is equal to or greater than the desired number of recovery paths.

According to another aspect of the present invention, in a computer, a method for establishing a work path and a recovery path from one node to another node in an MPLS network, each node on a path to be established, a link between each node, and each link A first procedure of identifying network availability information such as a bandwidth value between the network and a maximum bandwidth value of a recovery path between the respective links; A second procedure for selecting a link having a bandwidth value greater than the requested bandwidth value of the link; A third procedure of obtaining a shortest path between nodes selected in the second procedure and setting it as a work path, and transitioning to the first procedure if the job setting fails; A fourth step of selecting nodes and links except for links in which the sum of the bandwidth values between the link and the maximum bandwidth of the recovery path is smaller than the requested bandwidth value, except for the set work path, when the work path is set in the third procedure. step; A fifth procedure of obtaining a shortest path between nodes obtained in the fourth procedure and setting it as a recovery path; And a computer-readable recording medium having recorded thereon a program for executing a sixth procedure for transitioning to the fourth procedure if the recovery path is set in the fifth procedure, and if the recovery path fails. .

According to another aspect of the present invention, in a computer, a method for establishing a work path and a recovery path from one node to another node in an MPLS network, each node on a path to be established, a link between each node, and each link A first procedure of identifying network availability information such as a bandwidth value between the network and a maximum bandwidth value of a recovery path between the respective links; A second procedure for selecting a link having a bandwidth value greater than the requested bandwidth value of the link; A third procedure of obtaining a shortest path between nodes selected in the second procedure and setting it as a work path, and transitioning to the first procedure if the job setting fails; A fourth procedure of setting the number of recovery paths to an initial value "0" after the working path is established; A fifth procedure for selecting a node and a link excluding a link whose sum of the inter-link bandwidth value and the maximum bandwidth value of the recovery path is smaller than the requested bandwidth value except for the set work path; A sixth procedure of obtaining a shortest path between nodes obtained in the fifth procedure and setting it as a recovery path; A seventh procedure for transitioning to the fifth procedure if the recovery path setting fails, and increasing the number of the recovery paths by "1" when the recovery path setting is made; If the value increased in the seventh step is less than the desired number of recovery paths, the program transitions to the fifth step; A record carrier is provided.

According to the network recovery path establishment method according to the present method, the recovery paths share bandwidth with each other, thereby providing more network resources for the work path, thereby maintaining the high reliability of the services provided by the network while also increasing the efficiency of the network. It can increase.

In other words, the mechanism proposed by the inventors of the present invention presupposes a full recovery in consideration of an explicit path setting considering a protection switching model and traffic engineering for fast recovery. In addition, if required by the distribution method, the path is calculated at each source, that is, an edge node, and the calculated path is used as an input of Constrained-Based Routing for setting up a working path and a recovery path.

In the MPLS network, the problem of finding the path from the edge node to another edge node, or LSP, is that the direction having the set "V" of all the nodes of the MPLS network, the set "E" of the links, and the set "W" of bandwidth for each link is determined. It can be represented by the missing graph G = (V, E, W). In fact, since each link provides the same bandwidth in both directions, it can be assumed that G is a symmetric directional graph.

The path recovery mechanism according to the present invention calculates and calculates a working path and a recovery path at the source node "s" when it is required to set the LSP from one edge node "s" to another edge node "t" in an MPLS network. Set the path by the result. If the available bandwidth of each link is known in the current network state, the information about the input network can be assumed to be a directional graph with weights. Then, given the source node "s", the destination node "t", and the required bandwidth "b", the work path and recovery path from "s" to "t" are obtained while satisfying the bandwidth "b".

The present invention proposes two bandwidth sharing recovery path establishment mechanisms. Each of them is a mechanism for 1: 1 path recovery and a mechanism for 1: n path recovery.

Each link must store the maximum of the currently available bandwidth and the bandwidth currently allocated to the recovery path. Therefore, it can be assumed that a graph having two weights with the graph G = (V, E, AV, RT), where AV is the available bandwidth value of each link and RT is the maximum of the bandwidth allocated by each link to the recovery path. Indicates a value. Initially AV (i, j) has the total bandwidth value of that link of link (i, j), and RT (i, j) has "0".

2, an MPLS system functional block diagram including a recovery mechanism in accordance with the present invention is shown.

Referring to FIG. 2, a path calculation block 201 that calculates a work and recovery path calculates a work path and a recovery path based on network availability information from the routing protocol 202 when a user requests it. When the LDP 203 is requested for setting each calculated path, the LDP sets the requested path and informs the path calculation block whether the setting is successful. When the path calculation block 201 receives a response indicating that the path setting has failed, the path calculation block 201 calculates the path again and requests the LDP to set the path again. This process is repeated until the desired working and recovery paths are completely set. In addition, the LDP informs the forwarding engine 204 of the work path and the recovery path, and in the normal case, forwards the packet through the work path, and when a failure is detected in the corresponding path, the LDP requests the forwarding engine to forward the corresponding recovery path.

Referring to FIG. 3, a bandwidth sharing recovery path setting mechanism for 1: 1 path recovery in a path recovery method according to the present invention is illustrated.

The mechanism shown in FIG. 3 is a mechanism for setting one recovery path for one work path, and selecting the shortest path as the work path and setting the shortest path for the graph excluding the work path as the recovery path. .

Referring to the recovery method illustrated in FIG. 3, when a routing request having a bandwidth "b" from the user to the edge node "s" from the user to another edge node "t" is generated, the routing protocol block requests and receives the available information of the network. do. At this time, the information about the received network has a directional graph G = (V, E, AV, RT) having a weight (301). Where "V" is a set of nodes, "E" is a set of links, "AV" has an available bandwidth value for each link, and "RT" is a maximum bandwidth value of a recovery path for each link. If no recovery path has been set for the link, the initial value is zero.

In order to find the shortest path that allows bandwidth, only the edges with the " AV " value of the link greater than " b "

In the selected graph G '= (V, E'), the shortest path from s to t is found (303).

Ask the LDP to set the found path to WP (304).

Receive a response from the LDP as to whether to set the path (305).

If the path to the WP is successfully established, the process proceeds to step 307 of setting a recovery path. Otherwise, the process of calculating the working path by receiving available information of the network from the routing protocol again.

Except for the work path set from the graph G, a graph G "= (V", E ") except for the link" e "having AV (e) + RT (e) <b is constructed (307).

The shortest path from G "to" t "to" t "is found (308).

Ask the LDP to set the found path as the recovery path (AP) (309).

A response is received from the LDP as to whether to set the path (310).

If the path to the AP is successfully established, the process is terminated. If not, it repeats the process of receiving the network's available information from the routing protocol and calculating the recovery path.

Next, referring to FIG. 4, there is shown a bandwidth sharing recovery path establishment mechanism for 1: n path recovery, which is one of the path recovery methods according to the present invention.

The mechanism shown in FIG. 4 is a mechanism for setting one recovery path for one work path, selecting the shortest path as the working path, setting the shortest path to the recovery path 1 for the graph except the working path, In the rest of the graph, the shortest path is set as recovery path 2 and recovery path n.

Processes from reference numerals "401" to "406" in FIG. 4 are the same as those indicated by reference numerals "301" to "306" in FIG.

If the working path "WP" is set successfully, "407" initializes the value to "0" in the variable "k" to count the number of recovery paths.

In FIG. 4, the process indicated by reference numerals 408 to 411 is the same as the procedure indicated above by reference numeral 310 to 310.

If the recovery path "AP" is successfully set, the process proceeds to step "413". If not, the process of repeating the process of receiving the available information of the network from the routing protocol and calculating the recovery path is repeated.

If "AP" is successfully set, "k" is increased by one (413).

If the number of recovery paths required by "k" is smaller than n, the process of setting the recovery paths is repeated. End if greater than or equal to n.

5 shows a method of calculating a bandwidth value in one link by a path recovery mechanism according to the present invention.

As can be seen from Fig. 5, when "i" and "j" are nodes, the link (i, j) between both nodes is initially set to the available bandwidth AV (i, j) value, which is the maximum bandwidth value of the link. It has 0 as the maximum bandwidth RT (i, j) of the recovery path. The AV (i, j) value becomes the AV (i, j) -b value when setting the working path with bandwidth “b” (501), and the RT (i, j) <b when setting the recovery path with bandwidth “b”. AV (i, j) = AV (i, j)-(b-RT (i, j)), RT (i, j) = b, and when RT (i, j) ≥b, The "value" and "RT" value do not change (502).

In the example illustrated in FIG. 5, when a work path having a bandwidth of 3 is set, the AV (i, j) value is changed from 6 to AV (i, j) -3 = 3 (503). In case of setting a recovery path with a bandwidth of 3, since RT (i, j) <3, the AV (i, j) value changes from 6 to AV (i, j)-(3-2) = 6-1 = 5, The RT (i, j) value is 3 (504).

As described above, the path recovery method according to the present invention is a method that can efficiently use network resources in most cases where the probability of node or link failure is low.

In addition, the 1: 1 mechanism can normally perform path recovery unless the working paths of the recovery paths sharing the same link fail at the same time. In fact, the probability of two paths failing at the same time is very low since the square of the probability of one path failing.

And, the 1: n mechanism allows n-1 to send a message using one of the other recovery paths if the working paths of recovery paths sharing the same link fail at the same time. This method is somewhat less efficient than the 1: 1 mechanism, but has an advantage of maintaining the efficiency and reliability of the network properly by adjusting the size of n.

The above two mechanisms according to the present invention can increase the efficiency of the network, and maintain a high reliability level, compared to the conventional method that generally secures all the bandwidth for the recovery path.

The description so far is directed to the preferred embodiments for the understanding of the invention, and the invention is not limited thereto, and is intended to those skilled in the art without departing from the scope and spirit of the appended claims. It is obvious that various modifications and variations are possible.

Claims (7)

A routing protocol 202 that stores network availability information and routing information; A route calculation block 201 implemented in a source node and calculating a work path and a recovery path based on network availability information stored in the routing protocol 202 in response to a user request; An LDP 203 for setting a path in response to a path setting request from the path calculation block 201 and detecting a failure of a network path; And MPLS, characterized by using a path recovery mechanism through the forwarding engine 204 which receives the set work path and recovery path from the LDP and forwards the packet to a normal path among the work path or the recovery path through an instruction of the LDP. system, The method of claim 1, The path calculation block 201 configures a working path and a recovery path from the source node to the destination node within the available bandwidth of the link between the nodes in the path while knowing the available bandwidth of the current network state. MPLS system, characterized in that. The method of claim 2, MPLS system, characterized in that the bandwidth of each link is calculated in accordance with the following equation when setting the work path and the recovery path. [Calculation of Bandwidth of Link in Setting Work Path] AV (i, j) = AV (i, j)-b [Calculation of Bandwidth of Link in Setting Recovery Path] If RT (i, j) <b, {AV (i, j) = AV (i, j)-(b-RT (i, j)), RT (i, j) = b} (Where AV (i, j) is the available bandwidth value of the link between nodes " i " and " j ", and RT (i, j) is the maximum allocated to the recovery path of the link between nodes " i " and " j " Bandwidth value.) In the MPLS network to establish a work path and recovery path from one node to another node, A first step of identifying network availability information such as each node on a path to be established, a link between each node, a bandwidth value between each link, and a maximum bandwidth value of a recovery path between each link; Selecting a node having a bandwidth value larger than a requested bandwidth value of the link; A third step of obtaining a shortest path between the nodes selected in the second step and setting it as a work path, and transitioning to the first step if the job setting fails; A fourth step of selecting a node and a link excluding a link in which the sum of the inter-link bandwidth value and the maximum bandwidth value of the recovery path is smaller than the requested bandwidth value except for the set work path when the work path is set in the third step; step; A fifth step of obtaining the shortest path between the nodes obtained in the fourth step and setting it as a recovery path; And And a sixth step of transitioning to the fourth step if the recovery path is not set when the recovery path is set in the fifth step. The method of claim 4, wherein A seventh step of setting the number of recovery paths to an initial value "0" after the working path is set; An eighth step of increasing the number of recovery paths by "1" after the recovery path is established; And And a ninth step of transitioning to a fourth step if the value increased in the eighth step is smaller than the number of desired recovery paths and ending if the value is greater than or equal to the desired number of recovery paths. In the MPLS network to establish a work path and recovery path from one node to another node, A first procedure of identifying network availability information such as each node on a path to be established, a link between each node, a bandwidth value between each link, and a maximum bandwidth value of a recovery path between each link; A second procedure for selecting nodes having a bandwidth value greater than the requested bandwidth value of the link; A third procedure of obtaining a shortest path between nodes selected in the second procedure and setting it as a work path, and transitioning to the first procedure if the job setting fails; A fourth step of selecting a node and a link excluding a link in which the sum of the inter-link bandwidth value and the maximum bandwidth value of the recovery path is smaller than the requested bandwidth value except for the set work path when the work path is set in the third procedure; step; A fifth procedure of obtaining a shortest path between nodes obtained in the fourth procedure and setting it as a recovery path; And A computer-readable recording medium having recorded thereon a program for executing a sixth procedure for transitioning to the fourth procedure if the recovery path is set in the fifth procedure, and if the recovery path fails. In the MPLS network to establish a work path and recovery path from one node to another node, A first procedure of identifying network availability information such as each node on a path to be established, a link between each node, a bandwidth value between each link, and a maximum bandwidth value of a recovery path between each link; A second procedure for selecting nodes having a bandwidth value greater than the requested bandwidth value of the link; A third procedure of obtaining a shortest path between nodes selected in the second procedure and setting it as a work path, and transitioning to the first procedure if the job setting fails; A fourth procedure of setting the number of recovery paths to an initial value "0" after the working path is established; A fifth procedure for selecting a node and a link excluding a link whose sum of the inter-link bandwidth value and the maximum bandwidth value of the recovery path is smaller than the requested bandwidth value except for the set work path; A sixth procedure of obtaining a shortest path between nodes obtained in the fifth procedure and setting it as a recovery path; A seventh procedure for transitioning to the fifth procedure if the recovery path setting fails, and increasing the number of the recovery paths by "1" when the recovery path setting is made; A computer-readable program recording a program for executing an eighth procedure for transitioning to the fifth procedure if the increased value in the seventh procedure is smaller than the desired number of recovery paths and ending if the value is greater than or equal to the desired number of recovery paths. Record carrier.