JP5081861B2 - Route switching method in packet relay network - Google Patents

Description

  Relay network switching method for relaying communication between a plurality of packet networks, and a relay network that switches to a backup path without going through a relay network operation system based on a failure of a clock supplied to a communication device constituting the relay network Relates to the switching method.

  In recent years, it is expected to construct a relay network that can simultaneously accommodate a synchronous communication network and an asynchronous communication network in a carrier. Examples of synchronous communication networks include networks using ATM devices and STM devices and networks providing dedicated line services. Examples of the asynchronous communication network include an Ethernet device, an IP device, and other networks using devices that do not require network synchronization. In particular, Ethernet (registered trademark; hereinafter, registered trademark notation is omitted) was a network technology used in access networks close to users, such as in the home or in the company, but in recent years it has become a relay network that is the core of carriers. There is a demand to use it. This is a technology that has made widespread use of Ethernet, because the device development costs of routers and switch devices that make up the network have been reduced, and the operation and technical reliability have been well established. This is because a network can be constructed at a low cost by shifting to or by introducing a packet network into the core network.

  With the widespread use of Ethernet, standards and organizations such as ITU-T recommendation Y1731 (Non-patent document 1) and IEEE recommendation 802.1ag (Non-patent document 2) In order to improve the maintainability of these standards, the OAM function for Ethernet is standardized by ITU-T Recommendation G8031 (Non-patent Document 3) for the route switching (protection) method using the OAM function. Maintenance management related to communication services over Ethernet is possible.

  Conventionally, as a relay network, a highly reliable network composed of a digital clock supply device (DCS) and a synchronous communication device has been used to handle a synchronous network such as an STM or an ATM network. Transition to the packet network is taking place. However, it takes a certain period of time to complete the transition to the packet network. Therefore, there is a form in which the synchronous network and the packet network are accommodated using the existing relay network. In this case, the packet data is also transmitted through the relay network supplied with the clock.

ITU-T recommendation Y1731 IEEE Recommendation 802.1ag ITU-T recommendation G8031 / Y1342

  The synchronization system device needs to synchronize the network, and has a clock interface for receiving a clock with high accuracy from DCS (generally, 10 **-12: ** represents a power, and the same applies hereinafter), and a reception clock A high-accuracy network synchronization is maintained by reproducing and operating a clock synchronized with the network using a PLL or the like. However, when the clock interface fails, the synchronous system device generally generates a free-running clock by an operation such that the internal PLL is free-running and continues its operation. At this time, the host operating system that manages the devices of the synchronous network connected to the device recognizes the clock failure, and the maintainer replaces the clock interface. During this time, the synchronous system device operates with a free-running clock (generally about 10 **-6) that is less accurate than DCS, and also requires time for the maintainer to replace the clock interface. Therefore, packet data cannot be normally transmitted / received due to a phase fluctuation accompanying a synchronization (frequency) shift with another apparatus during this period, and a frame drop occurs. On the other hand, according to standard switching means (ITU-T recommendation G8031) of the carrier grade network, when a communication failure occurs on the communication path as a monitoring target, the failure status is confirmed and switched between the edge devices of the relay network. It is necessary to process for a certain period of time to complete the process, and there is a possibility that frame dropping may occur.

  In order to solve this problem, it is necessary to provide a high-accuracy clock source (generally an oscillator with an accuracy of about 10 **-10) and a buffer in the apparatus. However, this method is not an effective solution because the unit price of the apparatus is high. Since the traffic from the access network that accommodates each user is concentrated in the relay network, there is a characteristic that packets are gathered. When the switching time becomes long, the resources of the relay network are consumed due to packet discard and packet retransmission associated with it. . Therefore, even if a high-accuracy clock source is not provided, it is required to provide a network and apparatus configuration for avoiding these phenomena and a route switching method using them.

When a clock interface failure occurs, switch to a preset backup path (non-working path).
Specifically, it is a route switching method for a packet relay network that operates in synchronization with a clock, and is connected to a plurality of first communication devices that are connected to a clock asynchronous network and transmit / receive signal packets to / from a user. A signal received by an arbitrary first communication device on a packet relay network that operates in synchronization with a clock from the network synchronization device. A first route for relaying a packet via an arbitrary second communication device and a first communication device connected to the destination asynchronous network of the signal packet; and a first route different from the first communication device that has received the signal packet The second communication device and the first communication device connected to the packet destination asynchronous network, and the communication device in the first route detects a clock abnormality, Notification of abnormalities 1st packet is generated and transmitted to the first communication device connected to the destination asynchronous network, and the first communication device connected to the destination asynchronous network that received the first packet instructs path switching. When the second packet is generated and transmitted to the first communication device that has received the packet via the second route, the first communication device that has received the packet transmits the first communication device on the second route. Changed the route to forward received packets.

As a result, it is possible to provide a communication system construction method capable of switching to a standby system while continuing data communication with a free-running clock, avoiding packet loss and reducing device cost.

It is a network block diagram which shows the structural example of the communication network which implements the path | route switching method of this invention. Similarly, it is the block block diagram which showed the structural example of the relay network and the apparatus in detail. Similarly, it is an operation | movement sequence diagram which shows the operation example of a relay network. It is operation | movement explanatory drawing which shows the operation example of the relay network by the path | route switching method of this invention. It is a signal block diagram which shows the structural example of the OAM frame used with a relay network.

Hereinafter, embodiments of a route switching method for a packet relay network according to the present invention will be described with reference to the drawings.
FIG. 1 is a network configuration diagram showing a configuration example of a communication network that implements the route switching method of the present invention. The relay network 01 is a packet transfer network that connects the access networks 02 to each other and transfers packets, and is a network in which the edge device 11 and the relay device 21 are connected in a mesh shape as shown in FIG. Each edge device 11 is connected to a packet device 22 constituting the access network 02. The packet device 22 is connected to a packet terminal device 12 installed in a user network (not shown; a home LAN or a corporate network). A packet from the packet terminal 12 is transmitted to the packet terminal 12 of the access network 02 as a destination via the access network 02 and the relay network 01.

  As will be described later, in the relay network 01, a route for transferring a packet is provided between the edge device 11 and the relay device 21, and when an abnormality occurs in this route, the route in the network 01 is switched. It is a configuration.

  FIG. 2 is a block diagram showing in detail a configuration example of the relay network and devices used therein, and shows a detailed configuration example of the packet transfer network 01, the edge device 11, and the relay device 21.

  The edge devices 11a and 11b and the relay devices 21a and 21b are managed by the monitoring unit 202 of the operation control unit 201 for managing the entire relay network 01 and each device. The operation control unit 201 may be installed in any of the devices in the network, or may be an operation control device that is separate from each device. The edge devices 11 a and 11 b and the relay devices 21 a and 21 b are supplied with a clock for network synchronization with higher accuracy than the clock supply device (DCS) 30. When each device is normal, when the synchronization reference clock from the DCS is received by the clock I / F 204, the devices in the relay network 01 are synchronized by regenerating the clock with a PLL (not shown) in each device. It is the composition which takes.

Each of the relay device 21 and the edge device 11 is connected to the operation control unit 201, and based on this instruction, the control unit 200 that monitors and controls the device and the network, and interfaces between the devices and other networks. The I / F unit 203 for transmitting and receiving the packet, the SW unit 206 having a switch (routing) function for transferring the received packet to the destination device, and the synchronization reference clock from the DCS are received. The clock I / F unit 204 reproduces the clock with a PLL (not shown). Although it has been described that the clock I / F unit 204 includes a PLL (not shown), the transmitter 205 includes a transmitter 205, and is configured to self-run when the reception clock is cut or an abnormality occurs. Note that one side (side facing the outside) of the I / F unit 203 of the edge device 11 installed at the edge of the relay network 01 is connected to the packet device 22 of the access network 02.
FIG. 6 is a diagram for explaining an operation example when a failure occurs in an arbitrary device in the relay network 01 (the relay device 21a in this embodiment). Further, FIG. 3 is an operation sequence diagram showing an operation example of the relay network in the same case as FIG.
When the normal relay network 01 is operating, if the active path S100 is set in the relay device 21a and the non-active path S101 is set in the relay device 21b, a normal signal frame (via the relay network 01) is assumed. The packets transferred between the access networks 02 (or packet terminal devices 12), hereinafter referred to as normal frames) flow from the edge device 11a to the edge device 11b via the relay device 21a. Here, (1) when the clock I / F unit 204 of the relay device 21a fails or the abnormality of the reception clock is detected, (2) the clock I / F unit 204 of the relay device 21a that detects the failure self-runs. First, a switching frame is generated by the I / F unit 203, and (3) is transmitted to the I / F unit 203 of the opposing edge device 11b. After the switching frame is generated, even if a normal frame is received from the edge device 11a, a signal frame (hereinafter referred to as a free-running frame) is transmitted with the clock frequency shifted because the clock in the device is free-running. It will be. (4) Next, the I / F unit 203 of the edge device 11b generates a route control frame and transmits it to the edge device 11a via the relay device 21b. (5) The relay device 21b relays the route control frame to the edge device 11b. (6) When the I / F unit of the edge device 11b receives the route control frame, the signal route S100 is routed through the relay device 21a until now. The route is switched to the route S101 via the relay device 21b. That is, the non-active path is set for the relay apparatus 21a that has set the active path so far, and the path is switched using the path that was the non-active path as the active path until now. As a result, the normal frame from the edge device 11a is transferred to the edge device 11b via the relay device 21b.

  FIG. 4 is an operation explanatory diagram showing an operation example of the relay network by the above-described route switching method, and is a diagram showing the correlation between the number of packets passing through the relay network and the switching time. FIG. 5A shows the operation when the path is switched using the switching frame described above. According to the switching method of the present invention, since the switching frame is used, the switching is completed when the switching time Ta is 200 to 400 milliseconds. Therefore, since the time required for the self-running frame to be transmitted in the operational system is also sufficient for this switching time, the self-running accuracy of the oscillator 205 of the clock I / F unit 204 is about 10 **-6 and a normal crystal oscillator is used. Even if it is used, packet loss can be prevented from occurring by appropriately adjusting the amount of buffer memory provided in each I / F 203 by using a packet retransmission procedure or the like. With the existing method (b), when a clock failure occurs, the clock interface is manually replaced after the host operation control unit detects the failure, so switching time (exchange time) in units of hours (minutes) occurs. Therefore, packet loss may occur depending on the transmission speed and the frame length.

FIG. 5 is a signal configuration diagram showing a configuration example of a control frame (OAM frame) defined and used in the relay network 01. In the relay network of the present invention, path switching as described above is realized by defining and transmitting the following control frames between the edge devices 11 or between the edge device 11 and the relay device 21. The information necessary for controlling the relay network as described above is inserted in the message (packet) area for realizing each OAM function defined in ITU-T recommendation G8031.
Reference numeral 60 shows an example of the configuration of an OAM frame that transmits the clock failure information detected by the clock I / F unit 204 of the relay device 21a to the opposing edge device 11b. Specifically, the OAM frame (FDI frame) The clock failure information is inserted into the failure flag 61 of the OAM payload and sent from the I / F unit 203 of the relay device 21a to the I / F unit 203 of the edge device 11b.

  70 also shows a configuration example of an OAM frame that transmits the clock failure information detected by the clock I / F unit 204 of the relay device 21a to the facing edge device 11b. Specifically, the OAM frame (AIS frame) The clock failure information is inserted into the failure flag 71 of the OAM data information and sent from the I / F unit 203 of the relay device 21a to the I / F unit of the edge device 11b. Whether to use the FDI frame or the AIS may be selected based on the configuration of the relay network 01.

  80 shows an example of the configuration of an OAM frame that transmits the path control information generated by the I / F unit 203 of the edge device 11b to the opposite edge device 11a. Specifically, the OAM frame (APS frame) The path control information is inserted into the OAM data information and sent from the I / F unit 203 of the edge device 11b to the I / F unit 203 of the edge device 11a. Note that, as with other signal frames, the relay device 21 only sees this frame and transfers it to the destination edge device, and terminates this frame and does not perform control operations for path switching.

01 ... Relay network, 11 ... Edge device, 21 ... Relay device,
30 ... DCS, 60, 70, 80 ... OAM frame,
200: control unit, 201: operation control unit,
203 ... I / F unit, 204 ... Clock I / F unit, 205 ... Transmitter,
206... SW unit, S100... Operation path, S101.

Claims (4)

A route switching method for a packet relay network that operates in clock synchronization,
A plurality of first communication devices connected to a clock asynchronous network and transmitting / receiving signal packets to / from a user; and a plurality of second communication devices connecting the first communication devices to each other, the first and second communication devices Each of which is connected to a packet relay network that operates in synchronization with the clock from the network synchronization device, and the signal packet received by the arbitrary first communication device is transferred to the arbitrary second communication device and the destination asynchronous network of the signal packet. A first route relayed via the connected first communication device, and another second communication device connected to the packet destination asynchronous network from the first communication device receiving the signal packet A second route relaying via the first communication device,
When the first or second communication device on the first path detects a clock abnormality from the network synchronization device, a first packet for notifying the abnormality is generated and connected to the destination asynchronous network. The first communication device connected to the destination asynchronous network that has transmitted to the first communication device and received the first packet generates a second packet instructing route switching and passes the second packet through the second route. When the packet is transmitted to the first communication device that has received the packet, the first communication device that has received the packet switches the route so that the packet received by the first communication device is transferred on the second route. A method for switching the route of a packet relay network characterized by the above.   The first route is an active route that is used for transmission and reception of the signal packet during normal operation, the second route is a non-active route, and any of the active devices of the relay network receives the reception 2. The path switching method according to claim 1, wherein when a clock abnormality is detected, the non-operation path is switched to the operation path and communication is continued.   If any of the active devices in the relay network detects an abnormality in the reception clock, the outgoing device provided by the active device before switching to its own device until the switch from the non-active route to the active route is accommodated. 3. The path switching method according to claim 2, wherein the signal packet is transferred to the destination asynchronous network based on the self-running clock of the device.   4. The path switching method according to claim 1, wherein an OAM packet defined by ITU-T recommendation 8031 is used for the path switching.

Images