KR100248406B1 - Multi-layer restoration method with pmc(per-planned multi-chooser) in dcs mesh networks - Google Patents

Abstract

The present invention relates to a multi-layer recovery method for recovering a network failure by using a circuit distribution system (DCS) in a synchronous network. In the prior art, a line plan line recovery is performed. Although line plan path recovery is performed or dynamic path recovery is performed in a third manner, in the present invention, a line plan multi-selector recovery step is added to the line plan line recovery step and the dynamic path recovery step. If the recovery is not completed in the aftermath of another line failure while the recovery message is propagated to recover the failure, the sender recognizes the failure of the multi-line by sending a failure detection response message containing the failure to recover. Select the nodes connected by the failed node with the multi-selector. To a network failure and recovery through the other of the track, and the recovery for the untested channels by carrying out a recovery step wherein the dynamic path have a fast recovery time and a high recovery rate with respect to the multi-line and the node failure.

Description

Line Plan Multiple Selector Multi-Layer Recovery Method in Mesh Structure Network Using Circuit Distribution System

The present invention relates to a method for recovering from a failure of a network using a circuit distribution system (DCS) in a synchronous network, and more particularly, to self-healing for a failure of a network when configuring a network using a circuit distribution system. The present invention relates to a multi-layer recovery method that adds functions to increase network reliability and increase usability.

Currently developed transmission equipment processes large amounts of signals at high speed in high-speed, large-capacity networks, and the hardware is complicated, and the amount of software to control them is enormous, so that not only the failure of the equipment itself, but also the optical line cutting and natural disasters are caused. It is necessary to provide a stable service by recovering a network failure caused by a failure in a short time. 1 illustrates a self-healing network for recovering from network failures. As shown in the figure, the self-healing control method is divided into a central control method and a distributed control method (1), and is divided into a line and a path according to a signal recovery target (2). Algorithms that automatically and quickly recover for a long time are divided into pre-planned and dynamic methods (3).

As a restoration method mainly studied in the related art, there is a dynamic restoration method that performs restoration in real time after reflecting a network situation. However, this method has a disadvantage in that the recovery performance is reduced by increasing the message processing load by delivering a recovery message in the form of broadcasting by using a message communication channel (DCC: Data Communication Channel) between DCS nodes. On the other hand, the pre-planned restoration method shows high recovery performance by performing recovery through a predetermined restoration route according to the location of failure, but pre-recovering the recovery route for all failures of the network. It cannot be specified, and the amount of memory for this is enormous and vulnerable to network traffic fluctuations, so that no failure can be recovered successfully.

However, if you recover by systematically recovering these two recovery methods that have these characteristics in stages, you can get a constant recovery rate and fast recovery time obtained from the pre-planning. You can get a high recovery rate by performing a recovery on.

This is called a multi-layer restoration method. However, this method is effective for single line failures, but the advantages of the multi-layer recovery scheme are reduced by almost all restorations performed in the dynamic recovery stage for multiple line failures and node failures.

Multi-layer recovery is a network recovery method proposed by BT (British Telecom) in around 1992. While the existing network recovery methods are centered on one recovery method, there are two types of pre-recovery and dynamic recovery methods. The recovery method is applied step by step, and the structure is divided into a line and a path recovery method according to the recovery routing method. As a result, a new concept of network recovery that can highlight the advantages of each of the pre-planning and dynamic recovery methods and minimize the disadvantages can satisfy the failure recovery time and recovery rate.

In this scheme, if a network failure occurs, the system first attempts to recover using the pre-planned recovery method so that quick recovery is achieved. If the restoration is incomplete in the pre-planned recovery method, it is reinforced by dynamic recovery to satisfy the high recovery rate. Its recovery step structure is shown in FIG. The basic configuration of this recovery step is to recover using the recovery route database information according to the scenario planned in advance in preparation for the failure of the line, path, and node in the first recovery step (4), and complete recovery in the first recovery step. If is not performed, the dynamic line recovery is performed in the second recovery step (5). If the recovery is not successful even in the second recovery step, the recovery is terminated by performing dynamic path recovery in the third recovery step (6).

The multi-layer recovery method provides a good recovery performance for single line failure by adding the pre-planned recovery method to the existing dynamic recovery method, thereby providing a good recovery performance for a single line failure. There are some things that can be improved:

1) Problems caused by the change of network at the stage of planning

It is necessary to have a function that can effectively recover from the preplanning stage when network failures are delivered to each node in real time.

2) Database construction problem according to failure scenario in the preplanning stage

As the network expands and becomes more complex, not only does the amount of data storage to store recovery route information for failure recovery increase, but it also requires specific failure scenarios to recover from multi-line and node failures. The main plan restoration information, which is mainly composed, is to prepare for a single line failure.

3) Problems caused by multi-line failure and node failure

In case of multi-line failure and node failure, the recovery plan will be more likely to fail in the plan recovery phase, so almost all the recovery will be handled in the dynamic recovery phase. In addition, recovery message processing time increases due to broadcast of a recovery message, which is a disadvantage of the dynamic recovery method, and a problem arises that recovery time increases and recovery rate decreases due to competition for securing a spare channel with limited multiple failures.

As described above, the multi-layer recovery method has better recovery performance than other recovery methods for single-line failures, but multi-line failures and node failures cannot be recovered in the pre-plan recovery phase but are restored in the dynamic recovery phase. Failure to make full use of the benefits of recovery results in attenuation of recovery performance.

This problem reduces the maximum benefit of a multi-tiered recovery approach that provides fast recovery time. To solve this problem, it is necessary to reinforce the preplanning recovery step and add a new recovery step to cope with multi-line failure and node failure. Do.

Accordingly, the present invention proposes a recovery scheme that activates the preplanning recovery step by adding a pre-planned multi-chooser (PMC) recovery step to the preplanning step that can cope with multi-line failure and node failure. In the event of multiple line failures and node failures, the purpose is to improve the utilization of the preplanning step by performing the preplanning multi-selection recovery step instead of the existing preplanning recovery.

1 is a general classification of self-healing networks to which the present invention pertains.

2 is a block diagram illustrating a conventional multi-layer recovery method.

Figure 3a is a block diagram showing the procedure of the conventional multi-layer recovery method.

Figure 3b is a block diagram showing the procedure of the multi-layer hierarchical recovery method according to the present invention.

4 is a schematic diagram showing a process of selecting a multiple selector according to the present invention;

5 is a procedural diagram showing the beginning of a preplanned line recovery and PMC recovery phase.

6A is a structural diagram of a preplan recovery message according to the present invention;

6b is a structural diagram of a response message according to the present invention;

6c is a structural diagram of a preplanned multiple selector message according to the present invention;

7 is a network model diagram for simulating the present invention.

8 is a graph of cumulative recovery time and recovery rate in multi-line failures illustrating the differences between the present invention and the prior art.

9 is a graph of cumulative recovery time and recovery rate in node failures showing the difference between the present invention and the prior art;

* Description of the symbols for the main parts of the drawings *

10,40: Line planning line restoration stage 20: Dynamic line restoration stage

30: dynamic path recovery step 50: pre-plan path recovery step

100: pre-planner multiple selector recovery step 110: node A

120: C node 130: B node

In the PMC recovery step according to the present invention which achieves the above object, recovery is not completed in the aftermath of another line failure while each transmission node (DCS) propagates a recovery message to detect one line failure and recover its failure. If not, the existing multi-layer recovery method performs recovery in the next step, but if the recovery fails in accordance with the recovery message in the preliminary recovery step, a message containing a non-recoverable message is returned in its response message (NACK). By recognizing the failure of the multi-line, it is characterized by selecting the nodes connected to the node having the failure as a multi-selector to recover through the other line.

In other words, in order to cope with multi-line failure and node failure, PMC (Pre-planned Multi-Chooser) recovery method is added to the sequential steps of pre-planned line recovery, dynamic line recovery, and dynamic path recovery. It is.

As described above, this recovery method reduces the load of the data communication network due to an increase in recovery time generated when a multi-line failure and a node failure occur and broadcast message recovery propagation, among the disadvantages of the conventional multi-layer network recovery strategy.

The structure of the conventional multi-layer recovery scheme can be divided into a scheme focused on line planning and a scheme focused on dynamic recovery. Figure 3a shows a conventional scheme. As shown, the method focused on dynamic recovery consists of a line plan line recovery step (10), a dynamic line recovery step (20) inheriting the recovery rate, and a dynamic path recovery step (30) inheriting the recovery rate. For example, the scheme focused on the line plan includes a line plan track recovery step 40, a line plan path recovery step 50 that inherits its recovery rate, and a dynamic path recovery step 60 that inherits its recovery rate. On the other hand, in the present invention, as shown in FIG. 3B, the line plan multi-selector restoration step 100 is added between the line plan line recovery step 10 and the dynamic path recovery step 30, or the line plan line recovery step is performed. 40 and the dynamic recovery step 60.

The restoration method according to the present invention is an effective method for detecting and recovering multiple line failures and node failures by making full use of the line connection information with neighboring nodes possessed by each node in the line planning stage.

In the pre-plan recovery, both nodes that detect a line failure are selected as senders and selectors, and the sender delivers a recovery message through pre-specified recovery routes for the failing operating channels. This is done by verifying its usefulness and marking its status (ACK, NACK) in the response message and forwarding it to the sender via the same recovery route.

In the PMC recovery structure according to the present invention, the PMC recovery step is performed when multiple lines and node failures are detected. If a multi-line failure occurs between adjacent lines and cannot be recovered with a predetermined recovery route (when the sender receives a NACK response message), the UE transitions to the PMC recovery step as shown in FIG. 3B and performs recovery. That is, when the sender receives the NACK, the sender determines that the multi line failure or the node failure is selected, and the neighbor nodes connected to the second line failure node are selected by the multi selector. The sender node configures independent recovery routes with channels through each of the multiple selectors and the corresponding node failure, and starts recovery in the same manner as the multiple selectors and the preplanning recovery procedure. As described above, the rest of the channels except for the channel through which the recovery is performed through the PMC recovery step are performed by performing the dynamic path recovery, which is the third step recovery procedure of FIG.

Referring to Figure 4 will be described in detail the operation of the PMC according to the present invention.

① Multi-line fault detection

In line planning line recovery, network recovery is performed by setting one recovery route at the sender (node A) 110 of both adjacent nodes of the first line failure. The sender 140 transmits a recovery message through a predetermined recovery route to recover the channels of the failure line. Upon receiving this message, the intermediate node (C node) 120 determines whether the recovery route is useful by monitoring the status of the connected recovery route, and if it is normal, delivers the recovery message to the next node (node B) 130 to receive the message. To reach the selector as the destination. When the other line failure is detected in this transmission process, the node transmits a NACK response message to the sender through a predetermined NACK response route (150). If the recovery route is normal, the selector sends an ACK response message through the determined recovery route to complete the recovery. However, if the sender (node A) receives a NACK response message and reads, the line plan line recovery is stopped and the PMC recovery step is started.

② Multiple selector selection and recovery message delivery

As the first process of the PMC recovery step, neighboring nodes are selected as the multi-selector based on the node in which the second failure occurs, and the multi-selector PMC recovery message is transmitted to each selector (160). In this recovery message, a recovery route is configured for recovering channels through the second failed node of each selector, and the message is transmitted to multiple selectors through different recovery routes. In addition, for this recovery message, each selector sends an ACK or NACK response message to the sender to terminate the PMC recovery step, and for the unrecovered channel, the recovery is completed through a three-step recovery path dynamic path recovery.

③ PMC recovery procedure

In the basic plan recovery procedure, both nodes are divided into senders and selectors around the place where a failure occurs, and they are recovered through the following procedure. FIG. 5 shows a flowchart of a procedure showing the start of PMC recovery in the plan recovery phase. .

Start step 210: Detect faults and distinguish sender and selector nodes.

Sender step 220: Delivers a recovery message through a predetermined recovery route, and upon recovery of a NACK message, stops the line plan line recovery and transitions to the next step.

Transfer step 230: Intermediate nodes receiving the recovery message transmitted through the recovery route grasp the state of alarm and performance of the line included in the recovery route to be output, and deliver the recovery message to the next intermediate node. In case of detecting another failure, the NACK message is sent to the sender.

Chooser step 240: Finally, the node that receives the recovery message identifies the status of the recovery route and delivers an ACK / NACK message to the sender through the determined recovery route.

Response step 250: processing the ACK / NACK message, intermediate nodes on the recovery route changes the switch of the node when receiving the ACK message to recover the signal and forward it to the next node, the NACK message to the next node as it is To pass.

In the conventional multi-layer recovery method, when a sender receives a NACK message, the sender transitions to the next recovery step. However, in the PMC multi-layer recovery method according to the present invention, the sender decrypts the information by adding information indicating the cause of the NACK in the NACK message. It is recognized that it is a multi-line failure and the PMC recovery step is performed accordingly. In the PMC recovery phase, neighboring nodes are selected as multi-selectors around the second failed node, and these multi-selectors select a sender and a new recovery route for the channels delivered to the sender via the failed line. To perform.

6A and 6B show a message used in the pre-planning and PMC recovery steps of the restoration method according to the present invention. 6A is a structure of a pre-plan recovery message, and includes a sender and selector identifier, a lost channel capacity, and route information through which a recovery message is delivered, and an intermediate node provides a reserve channel capacity provided by a recovery route. Count hops and pass them to the selector. In addition, the response message is shown in Figure 6b, ACK and NACK can be displayed, and in the case of NACK it indicates the identifier of the faulted line and the capacity of the lost channel to its sender to the sender.

In addition, the structure of the PMC recovery message used in the PMC recovery step when detecting a multi-line failure is shown in FIG. 6C, and is composed of a sender, selected selector identifiers, respective capacity to be recovered, and a recovery route with each selector.

According to the invention,

-Can recover multi line failure and node failure.

-The load on the communication management network is reduced by processing fewer recovery messages than the dynamic recovery method.

-The recovery of reserve channels is reduced and high recovery rate can be obtained.

-Easy to locate fault

-Even in the case of multi-line failure, the restoration is performed only on the failed line without propagating the recovery message to the entire network, and does not affect the entire network.

In case of multi-line failure and node failure, recovery is not performed in the previous plan recovery phase in the existing multi-layer recovery method, but recovery is performed in the next step of dynamic recovery step, which increases the recovery time compared to the dynamic recovery method of the first step recovery method. Was done. However, with the addition of the PMC recovery step proposed in the present invention, the recovery can be performed in the pre-planned recovery step, thereby reducing the recovery time and obtaining a high recovery rate.

In order to verify the effect of the invention, a simulation was performed by modeling a network composed of DCS. As a tool for simulation, the model was programmed for performance analysis of the recovery processor using library functions provided by MIL3's OPNET. And the network environment for the simulation used the same network configuration applied for performance analysis in Bellcore (Bellcore) as shown in FIG. The network consists of 15 nodes and 28 lines.

The simulations were performed for multi-line failure and node failure, and the rate of change of the recovery time of the multi-layer recovery processor was analyzed using the number of hops and the service time of the recovery processor. Simulation models performed in the present invention are classified as shown in Table 1.

Classification of simulation model Simulation model Recovery steps Conventional Model M1 Line Planning-> Dynamic Track ---> Dynamic Track M2 Line Planning Line-> Line Planning Path ---> Dynamic Route Invention model M3 Line Planning Line-> Dynamic Line ---> Dynamic Route↘ Line Planning Multiple Selector↗ M4 Line Planning Line-> Line Planning Path ---> Dynamic Route↘ Line Planning Multiple Selector↗

1) Multi-line failure

As shown in FIG. 8, nine multi-line faults in which line faults occurred simultaneously on two adjacent lines were selected, and the recovery time and the recovery rate for each fault are individually accumulated. In multi-line failures, the recovery model (M3,4) with PMC recovery stage was better in recovery time and recovery rate than M1, M2. In particular, it was found that the PMC recovery step has a stable recovery performance by showing a constant recovery time for each failure, not proportional to the number of lost channels. The recovery rate was over 90% for almost all multi-line failures.

2) node failure

As shown in FIG. 9, seven nodes are selected and the results of failure of each node are individually shown by accumulating recovery time and recovery rate. The recovery time is shortened due to the high recovery rate in the PMC recovery phase and the low recovery in the dynamic recovery phase.In the M1 and M2 models, the recovery rate is dependent on the dynamic recovery phase. By recovering the error, the recovery time was greatly reduced. Therefore, the multi-layer recovery method with the proposed PMC recovery step is excellent for node failure. In addition, the restoration scheme focused on the preplanning was more advantageous in terms of the recovery performance than the restoration scheme focused on the dynamics.

Claims (5)

In the synchronous network of the mesh structure using the circuit distribution system, the first line recovery is performed to recover the network failure, and if it is not restored, the second recovery rate is inherited to the dynamic line recovery or line planning. In the multi-layer recovery method of performing a path recovery, or by performing the dynamic path recovery by inheriting the secondary recovery rate in the third When the transmission node detects one line failure and propagates a recovery message in order to recover the failure in the line plan line recovery step, when the restoration is not completed in the aftermath of another line failure, the failure details are included. By sending the error detection response message, the sender recognizes the failure of the multi-line, selects the nodes connected to the node with the failure as a multi-selector, recovers the failure of the network through other lines, and recovers the unrecovered channel. And a pre-planning multi-selector recovery step for inheriting the dynamic path recovery step. The method of claim 1, The pre-plan multi-selector recovery step, If a failure is detected during the propagation message of the pre-planned line recovery, multi-selection is recognized by adding information indicating the cause of the failure in the failure detection response message to recognize the multiple line failure, and multi-selecting neighboring nodes around the node where the failure occurred. Selecting a step; Delivering from the sender to each multiple selector a pre-planned multiple selector recovery message configured with a recovery route for recovering channels through the second failed node of each multiple selector; In response to the pre-selected multi-selector message, each selector performs a step of recovering a network failure by transmitting a NACK response message to the sender when the failure is detected, otherwise the sender sends an ACK response message to the sender. Multi-layer recovery method of pre-planner multiple selector in network of mesh structure using The method of claim 2, The recovery message of the line plan track recovery step, Comprising the sender and the selector identifier, the lost channel capacity and the route information to which the recovery message is delivered, the intermediate node calculates the reserved channel capacity and hop number provided in the recovery route and forwards them to the selector. Line Plan Multiple Selector Multi-Layer Recovery Method in Mesh-structured Network Using Circuit Distribution System. The method of claim 2, The response message of the line plan track recovery step, In the case of NACK, multiple line plan multi-selectors can be selected in the mesh structure network using a line distribution system, characterized by displaying the identifier of the failed line and the capacity of the lost channel as a property of the NACK. Tier recovery method. The method of claim 2, The pre-plan multi-selector recovery message, A method for multi-layer multi-layer recovery schemes in a mesh-structured network using a circuit distribution system, comprising a sender, selected selector identifiers, respective capacity to be recovered, and a recovery route with each selector.

Images