WO2006026930A1 - A method for optical meshwork service restoration - Google Patents

A method for optical meshwork service restoration Download PDF

Info

Publication number
WO2006026930A1
WO2006026930A1 PCT/CN2005/001445 CN2005001445W WO2006026930A1 WO 2006026930 A1 WO2006026930 A1 WO 2006026930A1 CN 2005001445 W CN2005001445 W CN 2005001445W WO 2006026930 A1 WO2006026930 A1 WO 2006026930A1
Authority
WO
WIPO (PCT)
Prior art keywords
network element
recovery
path
channel
distinguisher
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Ceased
Application number
PCT/CN2005/001445
Other languages
English (en)
French (fr)
Inventor
Yongliang Xu
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Huawei Technologies Co Ltd
Original Assignee
Huawei Technologies Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Huawei Technologies Co Ltd filed Critical Huawei Technologies Co Ltd
Priority to EP05783726.2A priority Critical patent/EP1737145B1/en
Priority to US11/568,766 priority patent/US7965938B2/en
Publication of WO2006026930A1 publication Critical patent/WO2006026930A1/zh
Anticipated expiration legal-status Critical
Ceased legal-status Critical Current

Links

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04J—MULTIPLEX COMMUNICATION
    • H04J14/00—Optical multiplex systems
    • H04J14/02—Wavelength-division multiplex systems
    • H04J14/0287—Protection in WDM systems
    • H04J14/0289—Optical multiplex section protection
    • H04J14/0291—Shared protection at the optical multiplex section (1:1, n:m)
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04J—MULTIPLEX COMMUNICATION
    • H04J14/00—Optical multiplex systems
    • H04J14/02—Wavelength-division multiplex systems
    • H04J14/0227—Operation, administration, maintenance or provisioning [OAMP] of WDM networks, e.g. media access, routing or wavelength allocation
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04J—MULTIPLEX COMMUNICATION
    • H04J14/00—Optical multiplex systems
    • H04J14/02—Wavelength-division multiplex systems
    • H04J14/0227—Operation, administration, maintenance or provisioning [OAMP] of WDM networks, e.g. media access, routing or wavelength allocation
    • H04J14/0241—Wavelength allocation for communications one-to-one, e.g. unicasting wavelengths
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04J—MULTIPLEX COMMUNICATION
    • H04J14/00—Optical multiplex systems
    • H04J14/02—Wavelength-division multiplex systems
    • H04J14/0278—WDM optical network architectures
    • H04J14/0284—WDM mesh architectures
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04J—MULTIPLEX COMMUNICATION
    • H04J14/00—Optical multiplex systems
    • H04J14/02—Wavelength-division multiplex systems
    • H04J14/0287—Protection in WDM systems
    • H04J14/0293—Optical channel protection
    • H04J14/0295—Shared protection at the optical channel (1:1, n:m)
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04J—MULTIPLEX COMMUNICATION
    • H04J14/00—Optical multiplex systems
    • H04J14/02—Wavelength-division multiplex systems
    • H04J14/0278—WDM optical network architectures
    • H04J14/0283—WDM ring architectures
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04J—MULTIPLEX COMMUNICATION
    • H04J14/00—Optical multiplex systems
    • H04J14/02—Wavelength-division multiplex systems
    • H04J14/0278—WDM optical network architectures
    • H04J14/0286—WDM hierarchical architectures
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04J—MULTIPLEX COMMUNICATION
    • H04J2203/00—Aspects of optical multiplex systems other than those covered by H04J14/05 and H04J14/07
    • H04J2203/0001—Provisions for broadband connections in integrated services digital network using frames of the Optical Transport Network [OTN] or using synchronous transfer mode [STM], e.g. SONET, SDH
    • H04J2203/0057—Operations, administration and maintenance [OAM]
    • H04J2203/006—Fault tolerance and recovery

Definitions

  • the present invention relates to a method for implementing optical network service recovery, and in particular, to a method for implementing service recovery of a grid optical network (hereinafter referred to as a grid).
  • GMPLS Control Plane Protocol - Multi-Protocol Label Switching
  • MPLS Control Plane Protocol - Multi-Protocol Label Switching
  • GMPLS is a family of IP-based protocols, including automatic discovery, routing, and signaling protocols. It serves as the control plane foundation for optical networks and supports automatic provisioning of optical connections and network failure recovery.
  • GMPLS introduces a new Link Management Protocol (LMP) to support automatic link discovery; the automatic discovery of links can obtain the connection relationship of adjacent NEs.
  • LMP Link Management Protocol
  • the information is forwarded to other networks in the network through routing protocols. Yuan released.
  • the link state is released by extending the "Open Shortest Path with Traffic Engineering with Priority Protocol (OSPF-TE)"; then, GMPLS extends the "Resource Reservation Protocol with Process Engineering (RSVP-TE)" or "" The Label Distribution Protocol (CR-LDP) for Restricted Routes establishes an end-to-end network connection.
  • OSPF-TE Open Shortest Path with Traffic Engineering with Priority Protocol
  • RSVP-TE Resource Reservation Protocol with Process Engineering
  • CR-LDP Label Distribution Protocol
  • the ITU-Telecommunication Standardization Sector incorporates the standardization of the optical network control plane into its Study Group 15 (SG15), which harmonizes the standardization of optical network control planes with "automatically commutated light”.
  • Network ASON mainly focuses on the standardization of requirements, frameworks and interfaces of the optical network control plane, and gradually forms a standard system based on ITU-T recommendation G.8080. It should be noted that the ASO standard itself does not include a protocol level implementation. ASON uses protocols from other standardization organizations, like the GMPLS protocol. As the basis of its agreement.
  • the traditional ring network protection mode such as shared multiplex section protection ring (MSPRing)
  • MSPRing shared multiplex section protection ring
  • the shared multiplex section protection ring requires that the capacity of the link it contains must be the same. If at least one link on the ring runs out of capacity, it needs to be expanded. There are two ways to expand the capacity: (1) The capacity of the upgrade ring (such as upgrading from STM-16 ring to STM-64). During the upgrade process, the existing services on the ring need to be migrated; (2) another ring is built.
  • the traffic of the original multiplex section can not be carried on the new ring, but the way of the ring stacking causes difficulties in maintenance and management.
  • the ring network has the problems of low bandwidth utilization and inconvenient capacity expansion, and cannot meet the transmission requirements of data services with rapidly changing traffic.
  • ASON can dynamically recover traffic affected by network faults.
  • the ASON network can support multiple fault recovery and provide higher service reliability.
  • the planning of the grid is directly driven by the service.
  • the corresponding link bandwidth can be planned according to the end-to-end traffic, which is more convenient and flexible than the ring network.
  • DWDM wavelength division technology
  • ASON networks can dynamically change the network topology according to the changes in business volume requirements, and adapt well to the needs of data services.
  • the ASON network based on the GMPLS control plane has the above flexible advantages, but the problem is that the recovery time of the rerouting by the resource reservation protocol (RSVP-TE) with the traffic engineering extension is in the second level, which cannot meet the requirements of the operator. It is even more difficult to adapt to the demand for voice services requiring 50ms protection time.
  • the problem of long recovery time in the grid affects operators' adoption of grid-based ASON networks.
  • CCAMP Common Control and Metrics Plane
  • RSVP-TE protocol extension supports end-to-end GMPLS-based recovery
  • the basic idea is to find a recovery path that is not related to the failure of the working path while establishing the working path of the connection.
  • the working path runs signaling, reserves resources and establishes cross-connections, and provides end-to-end service transmission capability.
  • the recovery path runs signaling, reserves resources but does not establish cross-connections, so resources on the recovery path can be shared and used.
  • the cross-connection establishment of its corresponding recovery path is triggered by signaling. Specifically, it is divided into two main steps: The first step is to reserve link resources in the network element on the recovery path, and the second step is to create a cross-connection between the network elements on the recovery path after the work path failure occurs to implement the recovery path. Activation.
  • the first step is to complete before the failure occurs, there is no real-time demand, and the completion speed of the second step affects the business interruption time, which has real-time requirements.
  • Both of the existing grid recovery schemes are driven by control plane message-based protocols (such as the GMPLS-based RSVP-TE protocol) rather than bit-based protocols like multiplex section ring protection.
  • a method for grid recovery comprising the steps of: a) reserving channel resources on a link through which a recovery path is passed and assigning a distinction to each channel; And determining an associated signaling channel for transmitting the differentiator; b) when detecting that the working path is a fault, transmitting, according to the associated signaling channel, a value indicating that the restored distinguisher activates the working path Restore the path.
  • the recovery path is activated according to the following steps: 1) the ingress network element of the recovery path searches for the next hop network element of the recovery path, and uses the associated signaling channel to send a value to the next hop network element to indicate recovery. a differentiator, and then establishing a cross-connection; 2) if the next hop network element is an intermediate network element of the recovery path, searching for the next next hop network element according to the received distinguisher value, to the next next hop network
  • the meta-delivery value is a discriminator indicating recovery, and then a cross-connection is established, until the egress network element receives the discriminator indicating the recovery, and proceeds to step 3). If the next-hop network element is the egress network element, the direct entry is performed.
  • the recovery path is activated according to the following steps: 1) the ingress network element of the recovery path searches for the next hop network element of the recovery path, establishes a cross-connection, and then uses the associated signaling channel to go down to the next hop network element.
  • the sending value is a discriminator indicating recovery; 2) if the next hop network element is an intermediate network element of the recovery path, searching for a next hop network element according to the received discriminator value, establishing a cross-connection, and then The next hop network element sends a value indicating the recovery of the distinguisher, until the egress network element receives the value indicating the recovery of the distinguisher, proceeds to step 3), and if the next hop network element is the egress network element, directly Go to step 3); 3) The egress network element establishes a cross-connection.
  • the information is the value of the distinguisher indicating the recovery.
  • the establishing the cross-connection means that the ingress network element is bridged in the forward direction of the connection, and the cross-connection is switched to the recovery path in the reverse direction of the connection; the intermediate network element establishes a bidirectional cross-connection; the egress network element crosses the handover in the forward direction of the connection to Restore the path and establish a bridge in the reverse of the connection.
  • the network element searches for the next network element by using a recovery mapping table.
  • the recovery mapping table may be calculated by the network management system or the planning system of the optical network or calculated by each network element by using a signaling; and may be set by the network maintenance personnel on a network-by-network basis or by assigning a distinguisher value to the recovery path. Let the process set.
  • said associated signaling channel is an unused byte of channel level overhead in a physical frame structure of said optical network.
  • the associated signaling channel is an unused overhead in the service layer of the physical frame structure of the optical network, and the transmitted distinguisher is unique within the network element port that sends the physical frame.
  • the associated signaling channel is an unused overhead in the service layer of the physical frame structure of the optical network, and the transmitted distinguisher establishes a corresponding relationship with the channel of the recovery path.
  • the associated signaling channel is an HDLC protocol signaling channel running on multiple overhead bytes.
  • the distinguisher employs bit compression coding to save bandwidth of the associated signaling channel.
  • the step b) detects the working state of the working path through the service layer alarm detection, or detects the working state of the working path through the channel layer alarm.
  • the service layer alarm includes signaling loss, frame loss, multiplex section alarm indication, multiplex section remote alarm indication, multiplex section signal degradation alarm, and signal loss of the OTN optical network of the SDH/SONET optical network, At least one type of alarm in the remote alarm indication;
  • the channel alarm includes an alarm indication of a channel/virtual path of the SDH/SONET optical network, a channel error violation limit, a remote defect indication, a remote fault indication alarm, and an OTN light At least one of a signal loss, a signal degradation, and a remote alarm indication of the network.
  • the step of the cross-connection on the recovery path includes: 1) the ingress network element switches the service to the receiving side to the working path, The idle distinguisher is sent to the intermediate network element along the recovery path; 2) after receiving the idle distinguisher, the intermediate network element searches for the channel according to the original distinguisher value of the recovery path, and sends the idle distinguisher to the channel; repeat step 2) until The egress network element receives the idle state distinguisher; 3) the egress network element completes the forward cross-connection switch of the bidirectional connection, and the two-way service is switched to the working path in both directions, and the bidirectional connection reverse bridge is removed; 4) The egress network element reversely transmits the idle state overhead along the recovery path.
  • the intermediate network element removes the cross-connection in both directions, and the ingress network element removes the bridging on the transmitting side.
  • the problem of fast recovery of the grid can be solved, the recovery speed of the grid is improved, and the existing equipment and the software therein are not required to be changed too much, and the conditions for the wide application of the grid recovery are created.
  • Figure 1 is a schematic flow diagram of one embodiment of the method of the present invention.
  • FIG. 2 is a schematic diagram of channel overhead of VC-3/VC-4/VC-4-XC in SDH
  • FIG. 3 is a schematic diagram of overhead in SDH STM-1 line frame format
  • Figure 4 shows the coding format of each channel distinguisher transmitted over the HDLC link on the STM-64 port in the SDH network
  • Figure 5 is a schematic diagram of an example of a grid
  • Figure 6 is a schematic diagram illustrating two ways of network recovery.
  • the term "grid” refers to a grid-like optical network.
  • the optical network includes a synchronous digital hierarchy (SDH), a synchronous optical network (SONET), and an optical transport network (OTN).
  • SDH synchronous digital hierarchy
  • SONET synchronous optical network
  • OTN optical transport network
  • "Bit-based protocol” refers to a protocol implemented by repeatedly transmitting a bit-coded signal sequence over a dedicated communication channel, which has the advantage of implementing a single cartridge, requiring less software support, and predictable transmission delay.
  • Message-based protocol refers to a protocol implemented by transmitting message packets, such as the Internet Protocol (IP), which is flexible in use, but requires more software support and poor delivery delay predictability.
  • IP Internet Protocol
  • Link refers to a fixed connection between two optical network elements, usually based on optical fibers or specific wavelengths therein. Multiple “channels” are combined by various multiplexing techniques and transmitted over a link, and at the opposite optical network element. Re-demultiplexed to restore the original multiple channels.
  • Working path refers to the network connection used to transfer customer signals between different locations.
  • Recovery path refers to the network connection used to recover the transmission of customer signals when the working path fails, preventing customer signals due to network failure.
  • the recovery path can only reserve channel resources, these resources can be shared by multiple recovery paths; the working path and the recovery path are usually bidirectional in the optical network, the establishment process usually from one end (ie “initiating end") to the other end (ie “end end”) in order, such that the one-way connection in the bidirectional connection that is consistent with the direction of establishment is hereinafter referred to as "forward", while the opposite one-way connection is called "reverse”
  • resource reservation refers to pre-reserving channels and cross-resources for the establishment of network connections to prevent the establishment of other connections.
  • Restoration path activation refers to the use of reserved resources to establish crossover on a predetermined recovery path.
  • “Differentiator” means that the upstream and downstream network elements on a particular recovery path are The value assigned and agreed upon on the link is used to indicate that the peer end uses the channel to activate the corresponding recovery path between the upstream and downstream network elements, and uniquely determines the recovery path when the multiple recovery path shares the reserved channel.
  • the code uniqueness of the distinguisher can be channel-level or link-level, and can be transmitted using bit-based protocols or through message-based protocols.
  • the method for recovering a grid of the present invention includes the steps of: a) reserving channel resources on a link through which the recovery path passes and assigning a discriminator to each channel; And determining an associated signaling channel for transmitting the differentiator; b) detecting an operating state of the working path; c) detecting, when the working path is faulty, sending a value through the associated signaling channel to indicate recovery A distinguisher that activates a recovery path of the working path.
  • the distinguisher of the present invention has two functions, the first is to indicate whether recovery of each channel of the recovery path is required, and the second is when the channel resources on the recovery path are shared, that is, when one channel of the link is used for multiple recovery paths , to determine which recovery path the channel resource is used for.
  • the discriminator is related to the shared use of the channel, and the number of bits encoded must take into account the maximum number of shares of the recovered resources allowed in the actual network. The more shares are shared, the higher the resource utilization, but the higher the risk of recovery failures when competing for resources during recovery, usually between 3 and 10 can get better results. Considering that regardless of whether a fault occurs or not, the overhead of the transport distinguisher (which will be described later) is always passed.
  • the distinguisher can pass a special value (such as 0) to indicate that the recovery channel is idle.
  • the distinguisher coding can be performed by bit compression coding, and each distinguisher occupies 2-4 bits, and each corresponding overhead byte can be divided into multiple distinguishers by bit division.
  • Multi-multiframe coding can further compress overhead. 8 Multiframe coding can compress the required number of overhead bytes to 1/8. The problem is that the overhead detection time is also extended by 8 times.
  • Example: An STM-64 link, considering extremes, all 64 channels are used for recovery. Each channel requires 4 bit transfers, and the number of overhead bytes required to transmit the differentiator is 64 x 4/8 32. If 8 multiframe coding is used, 7 bytes are used to transmit the distinguisher and 1 byte is used to transmit the other ⁇ byte check code, then the required number of overhead bytes is 32/7, ie 5 bytes is enough It is.
  • the discriminator also includes an encoding indicating a normal state, i.e., an encoding that does not require activation of the recovery path, such as zero. That is to say, when the network element receives the code 0 indicating the normal state, and does not change from other coded values to 0, no action is required.
  • FIG. 1 is a schematic diagram of the channel overhead of VC-3/VC-4/VC-4-XC in SDH.
  • the standard cascade (VC-4-Xc) type of Virtual Container-3 (VC-3) / Virtual Container-4 (VC-4) / Virtual Container-4 in SDH The channel overhead occupies 1 column for a total of 9 bytes, of which the 5-8 bits of the K3 byte are spare bits in the standard, and these 4 bits can be used to transmit the distinguisher; in addition, the N1 byte (network operator) Bytes) can also be used to pass a distinguisher.
  • the sending network element of the distinguisher needs to set the distinguisher value to the overhead byte of the corresponding channel of the corresponding link, and the downstream node detects the corresponding cost value to obtain a distinguisher, and obtains the port (link) and the channel identifier from the position of the overhead. .
  • Figure 3 shows the overhead in the SDH STM-1 line frame format, where the following 5 rows and 9 columns are complex With the overhead of the segment layer, among these overheads, "blank" is not used.
  • the overhead shown in Figure 3 is repeated 64 times in a row, that is, the position of each byte in the figure represents 64 bytes. When transmitting on the actual line, the 64 bytes are sent sequentially. Then 64 repetitions of the next overhead byte are transmitted.
  • the position of the overhead is usually expressed by (row, column, repeat count), such as (6, 1 , 1), that is, the sixth row and the first column are repeated for the first time, corresponding to D4 bytes.
  • the multiplex section overhead has been used, and the M1 overhead byte uses only the position of the 3rd repetition of its corresponding 64 repetitions, that is, the overhead position (9, 6, 3), other overheads only occupy the first repeated position, and other overhead locations are undefined. There are approximately 5,000 undefined overhead bytes for the STM-64 line.
  • the location of these overheads in the service layer overhead is established in advance with all the channels carried by the service layer, so that the port and channel identifiers can be directly obtained from the position of the overhead in the frame structure.
  • the multiplex section overhead (6, 1 , 1 ) of an STM-64 link corresponds to the first channel
  • the overhead (6, 1 , 2) corresponds to the second channel...
  • This correspondence requires both sides of the line.
  • the correspondence must be the same.
  • all optical network elements can be configured the same.
  • the sending network element needs to obtain the location of the overhead according to the corresponding relationship of the channel number corresponding to the distinguisher, set the corresponding cost value of the corresponding port to the distinguisher value, and receive the network element monitoring service layer overhead, according to the corresponding relationship mentioned above.
  • the overhead location gets the channel number.
  • the code of the distinguisher should be unique within the link port that transmits the distinguisher.
  • the overhead byte When the overhead byte is used as the dedicated signaling channel of the transmission distinguisher, the overhead byte is repeatedly transmitted in the SDH/SONET network at a rate of 8000 frames per second, and the transmission distinguisher is implemented by setting the overhead byte value. , such that the distinguisher is always transmitting repeatedly, and the peer network element is The change of the overhead value is monitored by the hardware, and the change is controlled by the processor to perform corresponding processing.
  • DCN Data communication network
  • signaling channels see ITU-T Recommendation G.7712
  • HDLC High Capacity Digital Cellular Network
  • D4-D12 Low-High Capacity Digital Cellular Network
  • SDH Serial Digital Network
  • the Advanced Data Link Control Protocol protocol is used to transmit the distinguisher.
  • the signaling channel is recommended to be dedicated. If the optical network element cannot support a large amount of undefined overhead processing due to hardware limitations, the overhead byte can be saved by using the HDLC protocol to transmit the distinguisher.
  • HDLC belongs to the link layer protocol. Since only the coding of this protocol is used here, no other processes are used, and the processing overhead is also small, and the distinguisher can be quickly transmitted.
  • Figure 4 shows the encoding format of each channel distinguisher transmitted on the STM-64 port through the HDLC link, where the hexadecimal value 7E is used to identify the start and end of the HDLC frame; the length field is used to indicate the subsequent distinguisher. The byte length of the sequence. If the classifier uses 4-bit encoding and 64 channels occupy 32 bytes, the length field is 32. The check code is used by the receiving end to check the integrity of the frame structure; the transmitting network element continuously transmits the same.
  • the distinguisher frame if the receiving network element detects that the frame is correct, sends an acknowledgement frame to the sending network element, and the sending network element can stop sending the distinguisher frame after receiving the acknowledgement frame, until the next time the distinguisher value is changed and then the sending is started.
  • the acknowledgment frame can be sent in combination with the distinguisher frame. If sent separately, the length field is 0, indicating that the content of the distinguisher is not included later.
  • the discriminator can be transmitted by any of the above methods, and the corresponding transmitting and receiving methods have been described in detail above, and are collectively referred to as transmitting the discriminator through the "associated signaling channel". It should be noted that the above associated signaling channels are bidirectional, and the network elements at both ends of the link can send a distinguisher to the peer through it.
  • the present invention first reserves resources for the recovery paths of the working path and the working path. To do this, you first need to determine the working path and recovery path.
  • the working path and recovery path can be obtained by centralized calculation (see literature: Wayne D. Grover, "Mesh-based survivable networks-Options and Strategies for Optical, MPLS, SONET, and ATM Networking", p311-313, PRENTICE HALL PTR, ISBN 0-13-494576-X, ), can also be obtained through distributed protocols (see: IETF draft, RSVP-TE Extensions in support of End-to-End GMPLS - based Recovery). Any method capable of obtaining a working path and a recovery path can be used to implement the present invention. Then, the recovery path for the working path and the working path is resource reserved on the link through which it passes, and a discriminator is specified for the recovery path.
  • Figure 5 shows an example of a grid, including network elements A, B, C, D, E, F, G, and H.
  • the connection relationship is given by a thick solid line in the figure, and the working path 1 passes through the network element A. , B and C, whose recovery path passes through A, D, E, and C; Work path 2 passes through the network element? , G, H, its recovery path through F, D, E, and H.
  • the links in the figure, the working paths of the network connections, and the recovery paths are all two-way connections.
  • the related NEs need to establish a recovery mapping table for all the recovery paths, which can be configured through the NMS or through the signaling protocol.
  • the signaling protocol is used as an example to illustrate the process of establishing a recovery mapping table.
  • the centralized network management configuration can be regarded as a centralized implementation of the process, and is not separately described.
  • the ingress network element A receives a resource reservation request message for initiating the recovery path 1 from the network management system, where the information includes:
  • the inbound NE A reserves the corresponding resource of the recovery path 1, that is, channel 1 of port 3, and sends the request information to the next hop D specified by the source route;
  • the network element D receives the request message, reserves the resource according to the information given by the source route, and sends the request to the next hop E;
  • the network element D receives the request message, reserves the resource according to the information given by the source route, and sends the request to the next hop C;
  • the network element C receives the request message and finds that it is the egress node of the recovery path. For the recovery path, the channel 1 and the discriminator value are assigned on port 3. The recovery map shown in Table 2 is created, and the recovery path channel to be allocated is created. And the discriminator value is sent to the network element E through the response message in the opposite direction of the request message;
  • the network element E receives the response message and obtains the channel and discriminator values assigned by the network element C. 001445
  • the complex path allocates channel 1 and the discriminator value on port 1, creates a recovery map entry as shown in the first item of Table 6, passes the assigned recovery path channel and the discriminator value in the opposite direction of the request message through the response message.
  • the network element D receives the response message, and obtains the channel and discriminator values assigned by the network element E.
  • the channel 1 and the discriminator value 1 are allocated on the port 1, and the recovery as shown in the first item of Table 5 is created.
  • the mapping entry, the assigned recovery path channel and the discriminator value are sent to the network element A in the opposite direction of the request message through the response message;
  • the network element A receives the response message, obtains the channel and discriminator values assigned by the network element D, creates a recovery mapping entry as shown in Table 1, and completes the signaling processing in the recovery path reservation phase.
  • the interpretation of the content in the recovery mapping table is related to the role of the network element on the recovery path, and is divided into three categories: entry, middle, and egress network element:
  • the role differentiation of network elements is for feature recovery paths. For different recovery paths, specific network elements may belong to different roles. And in the direction from the ingress node to the egress node, the outgoing value of the upstream network element must be the same as the ingress distinguisher value of the peer network element (downstream network element).
  • the following describes the recovery mapping table established at the shared node when the channel resources on the recovery path are shared.
  • the restoration path 1 and the restoration path 2 share resources on the link DE.
  • the recovery mapping table shown in Table 3 and Table 4 is formed at the network element H.
  • the recovery mapping table shown in Table 5 and Table 6 is formed at the network element D and the network element E.
  • the network element D and the network element E are Different recovery paths are assigned different distinguisher values, that is, for shared recovery path resources, different recovery is required.
  • the path assigns different discriminator values.
  • the above method of establishing a recovery mapping table is only illustrative and is applicable to the case where the distinguisher is bound to the channel.
  • the recovery mapping table of the intermediate network element is obtained according to the value of the recovery port and the input distinguisher.
  • the egress network element recovery mapping table obtains (in) the recovery channel, the working path (out) port and the channel according to the (in) recovery port and the distinguisher.
  • the remarks are not required.
  • the calculation of the recovery mapping table can be separated from the setting of the network element, and the calculation can be solved centrally by the network management system or the planning system, or can be solved by the signaling method.
  • the delivery can be configured on a per-station basis or by signaling. There may be centralized calculation but signaling configuration.
  • the associated signaling channel of the transmission distinguisher is also specified.
  • the establishment of the associated channel has been described in detail above, and therefore will not be mentioned here.
  • Activation of the working path can be implemented through the control plane signaling protocol, such as the RSVP-TE protocol that supports GMPLS extension.
  • the method of activating the working path is well known (see the literature: J.P. Lang, Y. Rekhter, D. Papadimitriou (Editor),
  • the dedicated signaling channel After the working path is activated, under normal circumstances (ie, when the working path is working normally), the dedicated signaling channel sends an idle distinguisher such as all zero code.
  • the network element (including all NEs that are the originating and ending ends of the recovery path) determines whether the working path is interrupted by alarm monitoring. If an alarm is found and the working path is interrupted, the recovery path is activated. When performing alarm monitoring, it is necessary to distinguish whether the working path is based on zone recovery or path recovery to determine the alarm points and types to be monitored.
  • segment recovery In the case of segment recovery, all segments (that is, links) on the working path need to detect the service layer alarm of the working path.
  • multiplex layer alarms including signaling loss (LOS), frame loss (LOF), multiplex section alarm indication (MS-AIS), and multiplex section remote alarm indication (MS-RDI).
  • LOS signaling loss
  • LEF frame loss
  • MS-AIS multiplex section alarm indication
  • MS-RDI multiplex section remote alarm indication
  • B2SD multiplex section signal degradation
  • OTN Alarms such as loss of signal (SF) and remote alarm indication (BRDI) are detected. If these alarms occur, local repair needs to be performed on the link where the alarm occurs. That is, the path between the upstream and downstream nodes of the link is bypassed to avoid the faulty link, because multiple working channel interruptions caused by the link failure can pass.
  • Multiple recovery paths are used for recovery. These recovery paths need to be activated by the steps described later.
  • the channel alarms of the ingress NE and the egress NE of the entire working path need to be detected.
  • the channel/virtual path alarm indication (AIS), channel error violation limit (B3SD), and remote defect indication (RDI) and Remote Fault Indication (REI) alarms for OTN, are signal loss (SF), signal degradation (SD), and remote alarm indication. The occurrence of these alarms indicates that the corresponding working path is interrupted, and the activation of the associated recovery path needs to be driven by the steps described later.
  • FIG. 6 is a schematic diagram of two ways to explain network recovery.
  • the working path of the service connection AF passes through the ABCF. If path recovery is adopted, the recovery path passes through the recovery path ADEF that does not intersect the working path node (or link); if segment recovery is used, the end-to-end
  • the path itself does not consider the problem of recovery, but considers recovery on the link itself, for example, the working channel on link BC can be recovered through multiple recovery paths: path BDEC and BADEC.
  • the "working path of the recovery path” representation refers to the corresponding path where the recovery path can be restored. That is to say, the recovery path of the segment and the recovery path of the path are not distinguished.
  • the process of activating the recovery path is: 1) the ingress network element of the recovery path searches for the next hop network element of the recovery path, and uses the associated signaling channel to send a value to the next hop network element to indicate recovery. a distinguisher, and then establish a cross-connection; 2) if the next hop network element is an intermediate network element of the recovery path, searching for the next next hop network element according to the received distinguisher value, to the next next hop. The network element sends a value indicating the recovery of the distinguisher, and then establishes a cross-connection, until the egress network element receives the value indicating the recovery, and proceeds to step 3).
  • next-hop network element is the egress network element, Go to step 3); 3)
  • the egress network element establishes a cross connection.
  • the egress network element sends a confirmation distinguisher on a network-by-network element along a path opposite to the recovery path.
  • the acknowledgment discriminator value is the same as the discriminator value of the forward activation recovery path described above.
  • each network element first sends a distinguisher, and then establishes a cross-connection.
  • the value is sent to the discriminator indicating that the next network element is activated.
  • the ingress network element When a cross-connection is established, the ingress network element is bridged in the forward direction (consistent with the direction of the process) (the signal is simultaneously sent to the working path and the recovery path), and in the reverse direction, the crossover is switched to the recovery path; the intermediate network element is established.
  • Two-way cross-connection; the egress network element is exactly the opposite of the ingress network element, and the cross-connection is switched to the recovery path in the forward direction of the connection, and the bridge is reversed in the reverse direction.
  • the network element A finds that the working path is interrupted by the alarm monitoring. The interrupt is generated and the recovery process is started.
  • NE A uses the ingress port and channel of working path 1 to find the recovery port from network element A's recovery mapping table (Table 1).
  • the recovery port is 3, the channel is 1, and the distinguisher value is 1.
  • the association port 3 channel 1 association is started.
  • the signaling channel sends a discriminator (value 1) to the peer network element D; because the network connection is bidirectional, in the forward direction (consistent with the direction of the activation process), it is necessary to establish the inbound port 1 channel 1 to the recovery port 1 channel 1 Cross-connection between the input signals on the working and recovery paths; in the reverse direction, the cross-connections of the working paths need to be removed to establish the intersection from the recovery port 3 channel 1 to the ingress port 1 channel 1;
  • the intermediate network element D detects the value of the distinguisher transmitted by the network element A through the associated signaling channel, and finds that the distinguisher changes from the idle state 0 to 1, and recovers the recovery map by using the distinguisher value (see Table 5).
  • the interface is 3, the channel is 1, and the discriminator value is 1.
  • the content of the recovery overhead setting overhead of the interface is set according to the value of the discriminator; the bidirectional cross-connection is completed between the inbound interface 1 channel 1 and the outbound interface 3 channel 1.
  • the initiating network element that is the next recovery path sends a discriminator that activates the recovery path to the peer network element E through the associated signaling channel.
  • the egress network element C detects the value of the discriminator transmitted by the associated signaling channel, and finds that the discriminator value is changed from the idle state 0 to 1, and the recovery map is searched through the inbound interface 3 channel 1 and the discriminator value 1.
  • the idle discriminator value 0 needs to be sent to the upstream and downstream of the recovery path.
  • the network element receives the discriminator value and finds that the discriminator value changes from non-zero to zero. According to the original distinguisher value, the port and the channel information, the established cross-connection needs to be removed, and the downstream node continues to send the idle distinguisher 0 to the egress network element along the recovery path, if the corresponding cross-connection exists. Need to be removed as above. After each network element removes the cross-connection, it also needs to send the idle distinguisher 0 to the upstream network element.
  • the network connection needs to return from the recovery path to the original working path, and then release the cross-connection on the recovery path. These resources can be used for recovery of other network failures.
  • the steps of releasing the cross-connection on the recovery path include: 1) the ingress network element switches the service from the receiving side to the working path, and sends the idle distinguisher to the intermediate network element along the recovery path; 2) after the intermediate network element receives the idle distinguisher, Find the channel, the idle distinguisher sent to the channel according to the original distinguisher value of the recovery path; repeat step 2) until the egress network element receives the idle state discriminator.
  • the egress network element completes the forward cross-connection switching of the bidirectional connection, and the two-way service is switched to the working path in both directions, and the bidirectional connection reverse bridge is removed; 4) the egress network element is reversely transmitted along the recovery path In the idle state, the intermediate NE removes the cross-connection in both directions, and the ingress NE removes the bridging on the transmitting side.
  • the fault on the working path 1 is duplicated, and the ingress network element A detects the alarm disappearing event, and waits for a waiting recovery time first;
  • step 3 If the ingress NE does not detect that the alarm occurs again during the waiting recovery time, it starts. Reply process (step 3), otherwise the process ends and the ingress network element still waits for the alarm to disappear;
  • the ingress network element A first switches the reverse direction (receiving side) of the bidirectional connection to the working path, and sends the idle discriminator to the network element D along the recovery path through the port 1 channel 1;
  • the network element D receives the value of the idle distinguisher, and finds that the distinguisher value of the channel 1 channel 1 changes from 1 to the idle state 0. According to the original distinguisher value 1 of the recovery path, the recovery map is obtained to obtain the port 3 channel 1, and The distinguisher of the channel is also set to the idle state;
  • the network element E repeats the action of the network element D, and sets the channel 2 channel 1 to set the distinguisher to the idle state;
  • the egress network element C receives the discriminator of the idle state, completes the forward cross-connection handover of the bidirectional connection, and implements the bidirectional service to switch to the working path in both directions, and removes the bidirectional connection and reverses the bridging;
  • the egress NE moves the idle state overhead along the recovery path.
  • the NEs E and D remove the cross-connections in the two directions.
  • the NE A removes the bidirectional connection forward bridging.
  • the differentiator delivery process and cross-connection establishment of their activation process require a merge process.
  • the associated signaling channel used by the differentiator delivery process of the activation process is based on the HDLC protocol or based on the overhead byte multi-frame coding, all the recovery paths of the same path need to be combined and transmitted as much as possible to improve the transmission efficiency;
  • the receiving side needs to delay the processing of sending, transmitting, and cross-connecting different channel distinguishers by delaying the transmission period of several frames to improve efficiency.
  • the network connection recovery time refers to the time between the occurrence of a work path failure and the restoration of the path activation to restore the network connection, that is, the network connection interruption time.
  • the length of recovery includes the following factors:
  • Alarm monitoring time which is the time from the occurrence of a network failure (such as the fiber is cut off) to the alarm monitoring;
  • the alarm notification time If the alarm occurs in the middle of the network connection, and the recovery action needs to be processed at the ingress node of the connection, the notification message may need to be transmitted through the DCN.
  • the GMPLS extended RSVP-TE supports notification through notification.
  • the ingress and egress nodes send fault notifications;
  • the recovery signaling is transmitted through a path (Path) message, usually about 500 bytes per message, and the Path message is an instance per channel.
  • the recovery signaling described in the present invention only has one channel. Or smaller;
  • Protocol processing time received from recovery signaling, completes decoding, updates state machine, forwards recovery signaling, this process requires communication between tasks and other work, and needs to occupy processor time; corresponding to GMPLS extended RSVP-
  • the TE signaling protocol processing is relatively complicated, and the recovery signaling processing described in the present invention is very simple;
  • the cross-connection configuration time the network element needs to configure the cross-connection after receiving the recovery signaling, and the cross-connection adjustment action takes time.
  • the method of starting the cross-connection after forwarding the recovery signaling is performed, so that the network on the recovery path is obtained. Parallelization of meta-joining actions
  • the present invention saves processing time and accelerates the recovery process in factors 3 and 4. If the GMPLS extended RSVP-TE is running on the 576 kbits/s DCN composed of the overhead bytes D4-12 of the SDH, the path information of the recovery signaling path of the 64 VC-4 channels transmitting the STM-64 takes 500. x 8 x 64/576000, about 440 milliseconds.
  • the recovery signaling method adopted by the present invention if the overhead byte transmission is directly used, theoretically completes the signaling transmission of all 64 channels in 125 microseconds, and the detection processing of the added network element should also be completed in about 1 ms.
  • the transmission time is 50 X 8/576000, which is about 1 millisecond.
  • the protocol processing overhead of the GMPLS extended RSVP-TE is quite complicated, in the case of a few milliseconds, and the protocol processing described in the present invention requires only a simple table lookup process, and the processing time can be ignored, and RSVP-TE is Restoring signaling by channel serial processing is even more expensive.
  • the IP DCN used by GMPLS is of a shared nature.
  • the link state refresh of various notification messages and routing protocols that need to be performed when the fault occurs increases the burden on the DCN, so that the recovery process based on GMPLS signaling exists in time. Large fluctuations, and the dedicated recovery signaling channel described in the present invention can avoid this problem and achieve stability and predictability of recovery time.

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Data Exchanges In Wide-Area Networks (AREA)
  • Small-Scale Networks (AREA)

Description

一种网格状光网络业务恢复的方法
技术领域 本发明涉及一种实现光网络业务恢复的方法， 尤其涉及一种实现网格 状光网络（以下称格网）业务恢复的方法。
背景技术
传统的光传送网 北美以外地区采用的同步数字体系 （SDH )和北 美的同步光网络（ SONET )一都是基于人工操作的业务供给方式， 保护方 式也是基于环网或者线路的固定保护方式。 这里所述业务指网络连接， 故 本文中光网络业务即指光网络连接。 随着以因特网 （Internet )为代表的数 据业务的高速发展， 对传送带宽的需求也不断膨胀， 这种人工供给的操作 模式和环网保护方式越来越不能适应业务需求。
在这种背景下， 因特网工程任务组（IETF )将原先用于分组交换网 中数据交换的控制平面协议一多协议标签交换 ( MPLS ) ~ 广到光网络 中， 称通用多协议标签交换 ( GMPLS )。 GMPLS 是一族基于 IP技术的 协议， 包括自动发现、 路由、 信令协议， 作为光网络的控制平面基础， 支 持光连接的自动供给和网络故障恢复。
GMPLS 引入了新的链路管理协议（LMP )来支持链路自动发现； 通 过链路自动发现能得到邻接网元的连接关系， 在此: 上， 通过路由协议 将这些信息向网络中的其他网元发布。 在 GMPLS 中， 通过扩展 "带流量 工程的开放最短路由优先协议（OSPF-TE )" 实现链路状态发布； 然后， GMPLS 通过扩展 "带流程工程的资源预留协议（RSVP-TE )" 或者 "受 限路由的标签分布协议（CR-LDP )" 建立支持端到端的网络连接。
国际电联-电信标准化部（ITU-T )将光网络控制平面的标准化工作 纳入到其下的第 15研究組（SG15 ),该研究组将光网络控制平面的标准化 工作统一在 "自动交换光网络（ASON )" 下。 ASON主要侧重光网絡控 制平面的需求、 框架、 接口方面的标准制订， 逐渐形成了以 ITU-T建议 G.8080为核心的标准体系。 需要指出的是 ASO 标准本身并不包括一个 协议层面的实现。 ASON采用其他标准化组织的协议， 像 GMPLS协议， 作为其协议基础。
传统的环网保护方式， 如共享复用段保护环 （MSPRing ) 可以提供 50ms 的业务恢复时间， 但是需要预留 50 % 的带宽用于保护， 带宽利用 率低。 环网更大的问题在于它引入的限制， 共享复用段保护环要求其包含 的链路的容量必须一致， 如果环上至少有一段链路的容量用尽， 就要进行 扩容处理。 扩容有两种方式： （1 )升级环的容量（如从 STM-16 环升级 到 STM-64 ), 在升级过程中， 环上的现有业务需要迁移处理； （2 )再另 建一个环， 将原来复用段环容不下的业务量承载在新环上， 但是这种环叠 环的方式造成维护和管理上的困难。 总之， 环网存在带宽利用率低、 扩容 不方便的问题， 不能适应业务量快速变化的数据业务的传送需求。
ASON通过引入控制平面， 网络故障影响的业务量可以通过重路由 实现动态恢复， 通过格网 （MESH )组网， ASON 网络可以支持多重故障 的恢复， 提供更高的业务可靠性。 格网的规划是直接由业务驱动的， 可以 根据端到端的业务量来规划相应的链路带宽， 与环网相比更方便、 灵活。 结合波分技术（DWDM )， ASON 网络可以根据业务量需求的变化动态改 变网络拓朴， 很好地适应数据业务的需求。 基于 GMPLS 控制平面的 ASON 网络具备上面灵活的优点，但问题是按目前通过带流量工程扩展的 资源预留协议（RSVP-TE ) 实现的重路由的恢复时间在秒级， 不能满足运 营商的需求， 更不能适应要求 50ms保护时间的话音业务的需求。格网现 存的恢复时间长的问题影响了运营商对基于格网组网的 ASON 网络的采 用。
IETF 的公共控制和度量平面 （CCAMP ) GMPLS 保护和恢复工作 组在 "RSVP-TE Extensions in support ofEnd-to-End GMPLS-based Recovery ( RSVP-TE协议扩展支持端到端的基于 GMPLS的恢复）"草案中， 为解 决 MESH恢复可靠性和速度问题， 给出了 "共享格网恢复（shared-mesh restoration )" 的机制。 基本的思路是在建立连接的工作路径时同时也找到 一条与工作路径故障不相关的恢复路径。 工作路径运行信令， 预留资源并 建立交叉连接， 提供端到端的业务传送能力； 而恢复路径运行信令， 预留 资源但不建立交叉连接， 这样恢复路径上的资源可以共享， 被用于保护多 条故障不相关的多条工作路径。 在工作路径故障时， 通过信令触发其对应 的恢复路径的交叉连接建立。 具体地， 其分为两个主要步骤： 第一步在恢 复路径上的网元中进行链路资源预留， 第二步工作路径故障发生后驱动恢 复路径上的网元建立交叉连接实现恢复路径的激活。 其中的第一步是故障 发生前完成的， 没有实时性需求， 而第二步的完成的快慢影响业务中断时 间， 是有实时需求的。 现有的格网恢复方案这两步都通过控制平面的基于 消息的协议（如基于 GMPLS扩展的 RSVP-TE协议）驱动完成， 而不是 类似复用段环保护的基于比特的协议完成的。 基于消息的协议需要更复杂 的软件支持， 需要操作系统多任务间的协调完成协议动作， 这些操作需要 占有比基于比特的协议更多的时间， 而且完成时间的抖动更大， 这使得第 二步恢复路径的建立的实时特性受到影响。因而，尽管由于资源是预留的， 恢复路径的建立相比不事先预留资源的动态恢复要短， 但是通过 IP协议 传送的信令要解决 50 - 200ms这样的恢复时间是很困难， 特别是在高容 量的链路故障引发大量路径恢复的情况下， 恢复时间难以确保。
发明内容
本发明的目的是提供一种格网业务恢复的方法， 用以克服现有技术的 网格恢复慢的缺点。
为实现本发明的目的， 根据本发明的一个方面， 提供了一种格网恢复 的方法， 包括以下步驟： a) 为恢复路径在其经过的链路上预留通道资源并 为各通道分配区分器， 以及确定传送所述区分器的关联信令信道； b)当检 测出所述工作路径为故障时， 根据所述关联信令通道传送的值为指示恢复 的区分器激活所述工作路径的恢复路径。
优选地， 按照下述步骤激活所述恢复路径： 1 )恢复路径的入口网元查 找恢复路径的下一跳网元、 利用所述关联信令通道向下一跳网元发送值为 指示恢复的区分器， 然后建立交叉连接； 2 )如果所述下一跳网元是恢复 路径的中间网元， 则根据所接收的区分器值查找再下一跳网元、 向所述再 下一跳网元发送值为指示恢复的区分器， 然后建立交叉连接， 直至出口网 元收到值为指示恢复的区分器， 进入步驟 3 ), 如果所述下一跳网元为出口 网元， 则直接进入步骤 3 ); 3 ) 所述出口网元建立交叉连接。 可选地， 按照下述步骤激活所述恢复路径： 1 )恢复路径的入口网元查 找恢复路径的下一跳网元、 建立交叉连接， 然后利用所述关联信令通道向 下一跳网元发送值为指示恢复的区分器； 2 )如果所述下一跳网元是恢复 路径的中间网元， 则根据所接收的区分器值查找再下一跳网元、 建立交叉 连接， 然后向所述再下一跳网元发送值为指示恢复的区分器， 直至出口网 元收到值为指示恢复的区分器， 进入步驟 3 ), 如果所述下一跳网元为出口 网元， 则直接进入步骤 3 ); 3 )所述出口网元建立交叉连接。
进一步包括， 所述出口网元沿所述恢复路径的相反方向发送确认恢复 路径完成的信息的步骤。
所述信息为所述指示恢复的区分器的值。
所述建立交叉连接是指入口网元在连接的正向进行桥接， 在连接的反 向将交叉切换到恢复路径； 中间网元建立双向交叉连接； 出口网元在连接 的正向将切换交叉到恢复路径， 在连接的反向建立桥接。
优选地， 所述网元通过恢复映射表查找所述下一网元。
所述恢复映射表可由所述光网络的网管系统或规划系统计算或通过信 令由各网元计算； 并可由网络维护人员逐网元设置或在为所述恢复路径分 配区分器值时通过信令过程设置。
优选地， 所述关联信令通道为所述光网络的物理帧结构中通道级开销 的未使用的字节。
优选地， 所述关联信令通道为所述光网络的物理帧结构中服务层的开 销中未使用的开销， 所传送的区分器在发送物理帧的网元端口内唯一。
优选地， 所述关联信令通道为所述光网络的物理帧结构中服务层的开 销中未使用的开销， 所传送的区分器与所述恢复路径的通道间建立一一对 应的关系。
优选地， 所述关联信令通道为运行在多个开销字节上的 HDLC协议信 令通道。
优选地，所述区分器采用比特位压缩编码以节省关联信令通道的带宽。 优选地，所述步驟 b )通过服务层告警检测来检测工作路径的工作状态， 或通过通道层告警来检测工作路径的工作状态。 进一步， 所述服务层告警包括 SDH/SONET光网絡的信令丟失、 帧丢 失、 复用段告警指示、 复用段远端告警指示、 复用段信号劣化告警、 及 OTN光网络的信号丢失、远端告警指示中的至少一种告警； 所述通道告警 包括 SDH/SONET光网络的通道 /虚路径的告警指示、 通道误码越限、 远 端缺陷指示和远端故障指示告警、 及 OTN光网络的信号丟失、 信号劣化 和远端告警指示中的至少一种告警。
进一步， 在所述工作路径修复后， 包括鋒放恢复路径上的交叉连接的 步骤， 所述释放恢复路径上的交叉连接的步骤包括： 1 )入口网元将业务 将接收侧切换到工作路径， 沿恢复路径向中间网元发送空闲区分器； 2 ) 中间网元接收到空闲区分器后， 根据恢复路径原来的区分器值查找出通 道， 向该通道发送的空闲区分器； 重复步骤 2 )直到出口网元收到空闲态 区分器； 3 ) 出口网元完成双向连接的正向的交叉连接切换， 实现双向业 务两个方向都切换到工作路径上， 拆除双向连接反向的桥接； 4 )所述出 口网元沿恢复路径反向传递空闲态开销， 中间网元拆除两个方向上的交叉 连接， 入口网元拆除发送侧的桥接。
利用本发明的方法，能够解决格网快速恢复问题，提高格网恢复速度， 并且不需要对现有设备和其中的软件做太大改变， 为格网恢复的广泛应用 了创造条件。
附图说明
图 1是本发明的方法的一个实施例的示意流程图；
图 2是 SDH 中的 VC-3/VC-4/VC-4-XC 的通道开销示意图； 图 3是 SDH STM-1线路帧格式中的开销示意图；
图 4是在 SDH 网络中 STM-64 端口上通过 HDLC链路传送各通 道区分器的编码格式；
图 5是格网的一个示例的示意图；
图 6是说明网络恢复的两种方式的示意图。
具体实施方式
下面结合附图， 详细说明本发明。 附图仅用于说明， 不是对本发明范 围的限制。 在本文中， 术语 "格网" 是指网格状光网络。 其中光网络包括同步数 字体系 （SDH )、 同步光网络（SONET )和光传送网 （OTN )。 "基于比特 的协议"是指通过专用的通信信道重复传递按比特编码的信号序列来实现 的协议， 其具有实现筒单， 需要的软件支持少， 传递时延可预测的优点。 "基于消息的协议"是指通过传送消息分组的方式实现的协议， 如因特网 协议（IP )， 这种协议使用灵活， 但是需要更多的软件支持， 传递时延可 预测性差。 "链路" 指两个光网元间的固定连接， 通常基于光纤或者其中 的特定波长， 多个 "通道" 通过各种复用技术合并后通过一条链路传送， 并在对端光网元中重新被解复用恢复原来的多个通道。 "工作路径" 是指 用于实现客户信号在不同地点间传送的网絡连接， "恢复路径" 是指用于 在工作路径故障时用于恢复客户信号的传送的网络连接， 防止客户信号因 为网絡故障而被中断传送， 为提高网络带宽利用率， 恢复路径可以只预留 通道资源， 这些资源可以被多个恢复路径共享； 工作路径和恢复路径在光 网絡中通常是双向的， 建立过程通常从一端（即 "发起端"）到另一端（即 "结束端"）顺序进行的， 这样双向连接中与建立方向一致的单向连接下 文称 "正向"， 而相反的单向连接则称 "反向，，。 "资源预留" 是指为网络 连接的建立预先保留通道和交叉资源， 防止其他连接的建立占用。 "恢复 路径激活"是指利用预留的资源在预定的恢复路径上建立交叉连接， 使得 该路径真正可以用于传送业务。 "区分器" 是指特定恢复路径上的上下游 网元为该路径在特定链路上分配并且协商一致的值， 用于上下游网元间指 示对端使用该通道激活对应的恢复路径， 并在多恢复路径共享该预留通道 时通过它唯一确定恢复路径， 区分器通过关联信令通道传递， 区分器的编 码唯一性可以是通道级的也可以是链路级的， 可以使用基于比特的协议传 递， 也可以通过基于消息的协议传递。
图 1是本发明的格网业务恢复方法的一个实施例的流程图。 如图 1所 示，在本发明的一个实施例中，本发明的格网恢复方法包括的步骤有： a) 为 恢复路径在其经过的链路上预留通道资源并为各通道分配区分器， 并确定 传送所述区分器的关联信令信道； b)检测工作路径的工作状态； c) 当检测 出所述工作路径故障时， 通过所述关联信令通道发送值为指示进行恢复的 区分器， 激活所述工作路径的恢复路径。 在后文将结合具体的示例对这些 步驟进行详细的说明。
本发明的区分器有两个作用， 首先是表明是否需要进行恢复路径的 各通道的恢复， 其次是在恢复路径上的通道资源被共享时， 即链路的一个 通道用于多条恢复路径时， 用以确定通道资源用于哪条恢复路径。 因而， 区分器与通道的共享使用有关， 其编码的位数要考虑实际网络中允许的恢 复资源的最大共享次数。 共享越多则资源利用率越高， 但是恢复时争夺资 源而出现恢复失败的风险也越高， 通常在 3 - 10之间选择就可以得到比 较好的效果。 考虑到无论故障是否发生， 传递区分器的开销（将在后文说 明）都一直在传递， 在网络正常状态下， 区分器可以传递一个特殊值（比 如说 0 )来表示恢复通道为空闲状态。 为节省开销字节的占用， 区分器的 编码可以采用比特位压缩编码，每个区分器占有 2-4 比特，相应的每个开 销字节可以按比特位分割传递多个区分器。 通过多复帧编码可以进一步压 缩开销占有， 8 复帧编码可以将需要的开销字节数压缩到 1/8, 问题是开 销检出时间也要相应延长 8倍。
举例： 一条 STM-64 的链路， 考虑极端情况， 所有 64通道都用于 恢复。 每个通道需要 4 比特位传送， 则需要的用于传送区分器的开销字 节数目为 64 x 4/8 = 32。 如果采用 8 复帧编码， 其中 7 字节用于传送区 分器， 1 字节用于传送其他 Ί 字节的校验码， 那么需要的开销字节数为 32/7, 即 5 字节就足够了。
另外， 除了上述的用于特定恢复路径使用的区分器编码， 区分器还要 包括表示正常状态的编码， 即无需激活恢复路径的编码， 例如 0。 也就是 说， 当网元接收到这种表示正常状态的编码 0时， 而且也不是从其他编码 值变为 0, 则无需作任何动作。
区分器有多种传送方法：
1、 通过通道级开销传送， 如 SDH 的高阶通道开销中未使用的字节， 这样监测区分器值同时得到相应的端口和通道。 图 2 是 SDH 中的 VC-3/VC-4/VC-4-XC 的通道开销示意图。 如图 2所示， 在 SDH 中 虚容 器 -3 ( VC-3 ) /虚容器 -4 ( VC-4 ) /虚容器 -4 的标准级联 ( VC-4-Xc ) 类型 的通道开销占用 1列共 9个字节，其中的 K3 字节的第 5-8 比特在标准中 为备用比特， 这 4个比特可以用于传送区分器； 另外， N1字节 （网絡操 作者字节）也可以用于传递区分器。 在区分器的发送网元需要将区分器值 设置到相应链路的相应通道的开销字节上， 下游节点检测相应的开销值得 到区分器， 从开销的位置得到端口 （链路）和通道标识。
2、 通过服务层的开销中未使用的开销传送， 如 SDH复用段中的未 使用开销， 图 3给出了 SDH STM-1线路帧格式中的开销， 其中的下面 5 行 9列是复用段层的开销， 这些开销中， "空白" 的就是没有使用的。 对 于 STM-64 的链路， 图 3给出的开销都连续重复 64次， 即图中每个字节 的位置代表 64个字节， 在实际线路上传送时， 这 64个字节顺序发送， 然 后再传送下一个开销字节的 64次重复。 开销的位置通常按（行， 列， 重 复计数）来表示， 如（6， 1 , 1 ) 即第 6行第 1列第 1次重复， 对应 D4 字节。 复用段开销中除了 B2字节对应的 64次重复都已经使用了， Ml开 销字节只使用了它对应的 64次重复中的第 3次重复的位置上， 即开销位 置（9, 6, 3 ), 其他的开销都只占用第一次重复的位置， 其他的开销位置 都是没有定义的。 对于 STM-64 的线路大约有近 5000个未定义的开销字 节。 事先将这些开销在服务层开销中的位置与该服务层承载的所有通道建 立——对应关系， 这样从开销在帧结构中的位置就可以直接得到端口和通 道标识。 如一个 STM-64链路的复用段开销 （6， 1 , 1 )对应第 1通道， 开销 （6， 1 , 2 )对应第 2通道 ...... 这种对应关系要求线路两侧的对应关 系必须一致， 简化起见， 可以将所有的光网元都配置为相同。 发送网元需 要用区分器对应的通道号根据上面说的对应关系得到开销的位置， 将相应 端口的对应开销值设置为区分器值， 接收网元监视服务层开销， 根据上面 说的对应关系从开销位置得到通道号。
3、 通过服务层的开销传送， 而且不采用固定的捆绑关系， 这时区分 器的编码要在传送该区分器的链路端口内唯一。
以上通过开销字节作为传送区分器的专用信令通道时， 开销字节在 SDH/SONET 网络中是以每秒 8000帧的速度重复发送的， 传送区分器是 以设置开销字节值来实现的， 这样区分器是一直在重复发送， 对端网元则 通过硬件监测开销值的变化， 并将变化通过中断的方式上艮控制处理器进 行相应的处理。
4、通过专用信令通道组成的数据通信网（ DCN )传送（可参见 ITU-T 建议 G.7712 ),如在 SDH 的复用段开销中利用 D4-D12 或者其他未用字 节运行 HDLC (高级数据链路控制协议） 协议来传送区分器， 为支持区 分器传送的实时性， 该信令通道建议专用。 如果光网元因为硬件的限制不 能支持大量的未定义开销处理时，通过使用 HDLC协议来传送区分器可以 节省开销字节。 HDLC属于链路层协议， 由于此处仅使用该协议的编码， 并不使用其他的流程，因而处理开销也很小，可以快速传送区分器。图 4给 出 STM-64端口上通过 HDLC链路传送各通道区分器的编码格式，其中 的十六进制值 7E用于标识 HDLC帧的开始和结束；长度域用于指示其后 的区分器序列的字节长度， 如果区分器采用 4 比特编码， 64个通道占用 32个字节， 则长度域为 32; 校验码用于接收端检查帧结构的完整性； 发 送网元连续重复发送该区分器帧， 接收网元如果检查发现帧正确则向发送 网元发送确认帧， 发送网元收到确认帧后可以停止发送区分器帧， 直到下 次区分器值改变再启动发送。 确认帧可以与区分器帧合并发送， 如果单独 发送则长度域为 0， 指示后面不包括区分器的内容。
区分器可以通过上面任何一种方式传送， 相应的发送和接收方法已经 在上面详细描述， 下面统称通过 "关联信令通道" 传送区分器。 需要说明 的是上面的关联信令通道都是双向的， 链路的两端网元都可以通过它向对 端发送区分器。
下面详细说明本发明的方法。 如上所述， 本发明首先为与工作路径和 工作路径的恢复路径进行资源预留。 为此， 首先需确定工作路径和恢复路 径。 工作路径、 恢复路径可通过集中计算得到 （可参见文献： Wayne D. Grover, "Mesh-based survivable networks-Options and Strategies for Optical, MPLS, SONET, and ATM Networking" , p311-313, PRENTICE HALL PTR, ISBN 0-13-494576-X, ), 也可以通过分布的协议得到（可参见： IETF draft, RSVP-TE Extensions in support of End-to-End GMPLS -based Recovery )。 只 要能够得到工作路径和恢复路径的方法都可以用于实现本发明。 然后， 为工作路径和工作路径的恢复路径在其经过的链路上进行资 源预留， 并为恢复路径指定区分器。
图 5示出了格网的一个示例， 其中包括网元 A、 B、 C、 D、 E、 F、 G 和 H, 连接关系在图中按粗实线给出， 工作路径 1 经过网元 A、 B和 C, 它的恢复路径经过 A、 D、 E和 C; 工作路径 2经过网元？、 G、 H, 它的 恢复路径经过 F、 D、 E和 H。 图中的链路、 网络连接的工作路径和恢复路 径都是双向连接。
在恢复路径资源预留阶段， 相关网元需要为所有的恢复路径建立恢 复映射表， 可以通过网管配置， 也可以通过信令协议实施。 下面以信令协 议为例说明恢复映射表的建立过程， 集中的网管配置方式可以看成该过程 的集中实现， 不再单独说明。
恢复路径 1预留阶段的信令过程包括下面的步骤：
1. 入口网元 A从网络管理系统收到发起恢复路径 1的资源预留请求 消息， 其中的信息包括：
a) 入口网元（A )、 端口 （1 )和通道（1 , 可选）
b) 出口网元（C )、 端口 （2 )和通道（1 , 可选）
c) 业务粒度 ( VC-4 )
d) 源路由， 按 "网元-出端口-通道" 给出： A-3-l、 D-3-1和 E-2-1
2. 入口网元 A预留恢复路径 1的相应资源， 即端口 3的通道 1，将请 求信息发往源路由指定的下一跳 D;
3. 网元 D收到请求消息， 按源路由给出的信息， 预留资源， 将请求 发送到下一跳 E;
4. 网元 D收到请求消息， 按源路由给出的信息， 预留资源， 将请求 发送到下一跳 C;
5. 网元 C收到请求消息， 发现是恢复路径的出口节点， 为恢复路径 在端口 3分配通道 1和区分器值 1、 创建如表 2 所示的恢复映射表、 将分 配的恢复路径通道和区分器值沿请求消息相反的方向通过响应消息发送 给网元 E;
6. 网元 E收到响应消息， 得到网元 C分配的通道和区分器值， 为恢 001445
复路径在端口 1上分配通道 1和区分器值 1、 创建如表 6的第一项所示的 恢复映射表项、 将分配的恢复路径通道和区分器值沿请求消息相反的方向 通过响应消息发送给网元 D;
7. 网元 D收到响应消息， 得到网元 E分配的通道和区分器值， 为恢 复路径在端口 1上分配通道 1和区分器值 1， 创建如表 5的第一项所示的 恢复映射表项， 将分配的恢复路径通道和区分器值沿请求消息相反的方向 通过响应消息发送给网元 A;
8. 网元 A收到响应消息， 得到网元 D分配的通道和区分器值， 创建 如表 1所示的恢复映射表项， 恢复路径预留阶段的信令处理完成。
表 1. 网元 A的恢复映射表
入端口 入通道 入区分器值 出端口 出通道 出区分器值 备注
1 1 - 3 1 1 恢复路径 1 表 2. 网元 C的恢复映射表
入端口 入通道 入区分器值 出端口 出通道 出区分器值 备注
3 1 1 2 1 - 恢复路径 1 表 3. 网元 F的恢复映射表
入端口 入通道 入区分器值 出端口 出通道 出区分器值 备注
1 1 - 2 1 1 恢复路径 2 表 4. 网元 H的恢复映射表
入端口 入通道 入区分器值 出端口 出通道 出区分器值 备注
2 1 1 3 1 - 恢复路径 2 表 5. 网元 D的恢复映射表
入端口 入通道 入区分器值 出端口 出通道 出区分器值 备注
1 1 1 3 1 1 恢复路径 1
2 1 1 3 1 2 恢复路径 2 表 6. 网元 E的恢复映射表
Figure imgf000014_0001
恢复映射表中的内容的解释与网元在恢复路径上的角色有关， 分入 口、 中间、 出口网元三种情况：
1、 恢复路径的入口网元， 其中的入端口和入通道（可选）指网络连 接的入口， 入区分器未使用， 出端口、 通道给出恢复通道的出口， 出区分 器则给出了需要通过关联信令通道传送的区分器值；
2、 恢复路径的中间网元， 其中的入端口、 通道和区分器值是通过关 联信令通道接收到的， 出端口、 通道给出恢复路径出口， 出区分器则给出 了需要通过专用信令通道需要传送的区分器值；
3、 恢复路径的出口网元， 其中的入端口、 通道和区分器值是通过关 联信令通道接收得到的， 出端口、 通道给出工作路径的出口， 其中的出区 分器没有使用。
需要注意的是网元的角色区分是针对特点恢复路径而言的， 对于不同 的恢复路径， 特定网元可能分属不同的角色。 并且沿从入口节点到出口节 点的方向， 上游网元的出区分器值必须与对端网元（下游网元）的入区分 器值相同。
下面说明恢复路径上的通道资源被共享时， 在共享节点处建立的恢复 映射表。在如图 5所示的格网中， 考虑到工作路径 1和工作路径 2在故障 方面不具有关联性， 恢复路径 1 和恢复路径 2在链路 D-E上共享资源。 在为经过网元 A、 D、 E、 C的恢复路径建立了恢复映射表， 和为经过网元 F、 D、 E、 H的恢复路径建立恢复映射表之后， 则在网元？、 网元 H处形 成表 3、 表 4所示的恢复映射表， 在网元 D、 网元 E处形成了如表 5和表 6所示的恢复映射表， 网元 D和网元 E为不同的恢复路径分配了不同的区 分器值， 也就是说， 对被共享的恢复路径资源来说， 需要针对不同的恢复 路径分配不同的区分器值。
应注意， 上述建立恢复映射表的方法只是示意性的， 适用于区分器 与通道进行了绑定的情况。 本领域的技术人员应该意识到， 在未进行这种 绑定时， 网元恢复映射表的建立过程需要做相应修改： 中间网元的恢复映 射表才艮据入恢复端口和入区分器值得到入恢复通道、 出恢复端口、 通道和 区分器， 出口网元恢复映射表根据 (入）恢复端口和区分器得到 （入）恢 复通道、 工作路径（出）端口和通道。 另外， 备注部分也不是必须的。
另外， 恢复映射表的计算和在网元进行设置可以分开， 计算可以由网 管系统或规划系统集中解决， 也可以通过信令的方法分布解决。 而下发可 以逐站配置也可以用信令解决， 可能存在集中计算但是用信令下发配置的 情况
在资源预留的阶段， 还要指定传送区分器的关联信令通道， 关联信 道的建立已在前面详细叙述， 因而这里不再赞述。
随后， 激活工作路径。 工作路径的激活可以通过控制平面的信令协 议实施， 如支持 GMPLS 扩展的 RSVP-TE协议。 激活工作路径的方法 是公知的 （可参见文献： J.P. Lang, Y. Rekhter, D. Papadimitriou(Editor),
"RSVP-TE Extensions in support of End-to-End GMPLS-based Recovery," Internet Draft, Work in progress, draft -ietf -ccamp 一 gmpls -recovery -e2e -siganling -01.txt, May 2004. ), 因而本文不予赘述。
工作路径激活之后， 在正常情况下 （即工作路径正常工作时）， 该专 用信令通道发送空闲区分器如全零编码。 网元（包括所有作为恢复路径发 起端和结束端的网元）通过告警监视来判断工作路径是否中断， 如发现告 警， 并判断工作路径中断， 则进行恢复路径激活。 在进行告警监视时， 需 要区分工作路径是按区段恢复还是路径恢复来确定需要监测的告警点和 类型。
区段恢复情况下需要在工作路径上所有的区段（就是链路）检测工作 路径的服务层告警。 对于 SDH/SONET 需要监测复用段层的告警， 包括 信令丟失（LOS )、 帧丢失（LOF )、 复用段告警指示（MS-AIS )、 复用段 远端告警指示 （MS-RDI )、 复用段信号劣化 （B2SD )等， 对于 OTN 需 要检测信号丟失（SF )、 远端告警指示（BRDI )等告警。 如果这些告警发 生， 需要在告警发生的链路上作本地修复， 即在该链路的上下游节点间另 择路径迂回避开故障链路， 因为链路故障引起的多个工作通道中断可以通 过多条恢复路径实现恢复， 这些恢复路径都需要通过后文所述的步骤驱动 实现激活。
路径恢复下需要检测整个工作路径的入口网元和出口网元的通道告 警， 对于 SDH/SONET是通道 /虚路径的告警指示 （AIS )、 通道误码越限 ( B3SD )、远端缺陷指示（RDI )和远端故障指示（REI )告警，对于 OTN 则是信号丟失（SF )、 信号劣化（SD )和远端告警指示。 这些告警发生则 表示相应的工作路径中断， 需要通过后面所述的步骤驱动关联的恢复路径 的激活。
这里所说的区段恢复和路径恢复是网络恢复的两种方式。 图 6是用 于解释网络恢复的两种方式的示意图。 如图 6所示， 业务连接 A-F 的工 作路径通过经过 A-B-C-F, 如果采用路径恢复， 恢复路径经过与工作路径 节点（或者链路）不相交的恢复路径 A-D-E-F; 如果采用区段恢复， 则端 到端的路径本身并不考虑恢复的问题， 而是在其经过的链路本身考虑恢 复，如链路 B-C 上的工作通道可以通过多个恢复路径恢复：路径 B-D-E-C 和 B-A-D-E-C。
应该注意的是， 在本文中， 尽管为判断工作路径的工作状态区分了 区段恢复和路径恢复， 但本发明的方法可适用于这两种恢复。 "恢复路径 的工作路径" 表述是指恢复路径可实现恢复的相对应的路径。 也就是说， 并不区分区段的恢复路径和路径的恢复路径。
下面详细介绍激活恢复路径的过程。 在本发明的一个实施例中， 激活 恢复路径的过程为： 1 )恢复路径的入口网元查找恢复路径的下一跳网元、 利用关联信令通道向下一跳网元发送值为指示恢复的区分器， 然后建立交 叉连接； 2 )如果所述下一跳网元是恢复路径的中间网元， 则根据所接收 的区分器值查找再下一跳网元、 向所述再下一跳网元发送值为指示恢复的 区分器， 然后建立交叉连接， 直至出口网元收到值为指示恢复的区分器， 进入步驟 3 )， 如果所述下一跳网元为出口网元， 则直接进入步骤 3 ); 步骤 3 ), 所述出口网元建立交叉连接。 最后， 优选地， 出口网元沿与恢复路径 相反的路径逐网元发送确认区分器。 所述确认区分器值与上面说的正向激 活恢复路径的区分器值相同。
在上述过程中， 各网元先发送区分器， 然后才建立交叉连接。 当然， 可选地， 也可以在建立交叉连接之后， 才发送值为指示下一网元进行激活 的区分器。
建立交叉连接时， 入口网元在连接正向 （与激泠过程方向一致的）进 行桥接 (信号同时发送到工作路径和恢复路径）， 在反向则将交叉切换到 恢复路径；中间网元建立双向交叉连接； 出口网元则与入口网元正好相反， 在连接的正向将交叉切换到恢复路径， 反向则进行桥接
下面以图 5中恢复路径 1的激活为例，具体说明恢复路径激活的步骤：
1. 网元 A通过告警监视发现工作路径 1发生中断， 启动恢复过程；
2. 网元 A用工作路径 1的入端口和通道从网元 A的恢复映射表 (表 1 ) 查找得到恢复端口为 3, 通道为 1 , 区分器值为 1; 启动端口 3通道 1 的关联信令通道向对端网元 D发送区分器（值为 1 ); 因为网絡连接是双 向的， 在正向 （与激活过程方向一致的）需要建立入端口 1通道 1到恢复 端口 1通道 1之间的交叉连接，实现输入信号在工作、恢复路径上的桥接； 在反向， 则需要拆除工作路径的交叉连接建立从恢复端口 3通道 1到入端 口 1通道 1的交叉；
3. 中间网元 D检测网元 A通过关联信令通道传送的区分器的值， 发 现区分器从空闲态 0改变为 1 , 通过该区分器值查找恢复映射表（见表 5 ) 得到出恢复接口为 3, 通道为 1， 和出区分器值为 1 , 按出区分器值设置该 接口的恢复开销设置开销的内容； 在入接口 1通道 1和出接口 3通道 1之 间完成双向交叉连接； 并作为下一段恢复路径的发起网元通过关联信令通 道向对端网元 E发送指令激活恢复路径的区分器。
4. 后续的中间网元 E重复步骤 3直到出口网元；
5. 出口网元 C检测关联信令通道传送的区分器的值， 发现区分器值 从空闲态 0改变为 1 , 通过入接口 3通道 1和区分器值 1查找恢复映射表
(见表 2 )得到出端口 2通道 1 , 根据这些信息完成恢复路径出口的交叉 切换完成业务从工作路径到恢复路径的切换： 在正向， 需要拆除工作路径 的交叉连接， 建立从入端口 3通道 1到出端口 2通道 1的交叉； 在反向， 则需要增加出端口 2通道 1到入端口 3通道 1的交叉， 完成桥接；
6. 从出口网元 C 沿恢复路径相反的方法逐站回送区分器值， 直到入 口网元 A, 确认恢复路径完成。
通过以上步骤完成恢复路径的激活， 完成业务从工作路径到恢复路径 的恢复工作。 这些步骤只需要根据内存中的信息完成， 可以保证恢复路径 建立的快速性， 达到与共享复用段环 ASP协议类似的恢复速度。
在恢复路径的激活过程中的任何一步驟发生错误， 则需要向恢复路径 的上游和下游发送空闲区分器值 0， 网元收到该区分器值， 发现区分器值 从非 0到 0的转变， 需要根据原来的区分器值和端口、 通道信息查找得到 需要拆除的已建立的交叉连接， 下游的节点则继续沿恢复路径向下游发送 空闲区分器 0直到出口网元， 如果相应交叉连接存在也需要按上面的方法 拆除。 各网元拆除交叉连接后还需要向上游网元发送空闲区分器 0。
当工作路径上的故障修复后， 网絡连接需要从恢复路径回到原先的 工作路径， 然后释放恢复路径上的交叉连接， 这些资源可以用于其他网络 故障的恢复使用。
释放恢复路径上的交叉连接的步骤包括： 1 )入口网元将业务从接收 侧切换到工作路径， 沿恢复路径向中间网元发送空闲区分器； 2 ) 中间网 元接收到空闲区分器后， 根据恢复路径原来的区分器值查找出通道， 向该 通道发送的空闲区分器；重复步骤 2 )直到出口网元收到空闲态区分器。 3 ) 出口网元完成双向连接的正向的交叉连接切换， 实现双向业务两个方向都 切换到工作路径上， 拆除双向连接反向的桥接； 4 )所述出口网元沿恢复 路径反向传递空闲态开销， 中间网元拆除两个方向上的交叉连接， 入口网 元拆除发送侧的桥接。
以工作路径 1为例说明恢复过程的步骤：
1. 工作路径 1上的故障被^ i 复，入口网元 A监测到告警消失事件， 首 先等待一个等待恢复时间；
2. 如果在等待恢复时间内入口网元 A未监测到告警再次发生，则启动 回复过程（步骤 3 )，否则该过程结束，入口网元仍然等待告警消失；
3. 入口网元 A将双向连接的反向（接收侧）首先切换到工作路径， 沿 恢复路径通过端口 1通道 1向网元 D发送空闲区分器；
4. 网元 D接收到空闲区分器值， 发现端口 1通道 1的区分器值从 1 变为空闲状态 0, 根据恢复路径原来的区分器值 1查找恢复映射表 得到出端口 3通道 1 , 将该通道的区分器也设置为空闲态；
5. 网元 E重复网元 D的动作，将出端口 2通道 1设置区分器为空闲态；
6. 出口网元 C收到空闲态的区分器，完成双向连接的正向的交叉连接 切换， 实现双向业务两个方向都切换到工作路径上 , 拆除双向连接 反向的桥接；
7. 出口网元 C沿恢复路径反向传递空闲态开销， 网元 E、 D拆除两个 方向上的交叉连接， 网元 A拆除双向连接正向的桥接。
为了提高恢复速度， 如果多条恢复路径经过的路径相同， 则它们的激 活过程的区分器传递过程和交叉连接建立需要合并处理。 如果激活过程的 区分器传递过程使用的关联信令通道是基于 HDLC协议的或者基于开销 字节多复帧编码的， 则发送时需要将所有同路径的恢复路径尽可能合并发 送， 提高传送效率； 对应其他的基于开销字节关联信令通道的， 则接收侧 需要通过延迟几个帧发送周期的方法， 实现不同通道区分器接收、 发送和 交叉连接的合并处理， 提高效率。
网络连接恢复时间指从工作路径故障发生到恢复路径激活恢复网络 连接之间的时间， 也就是网络连接中断时间。 恢复时间的长短包括下面的 因素：
1. 告警监测时间， 即从网络故障发生（如光纤被切断）到告警监测发 现的时间；
2. 告警通知时间，如果告警发生在网络连接的中向， 而恢复动作需要 在连接的入口节点处理的话则可能需要通过 DCN传递通知消息， 如 GMPLS 扩展的 RSVP-TE 支持通过通知 （Notification ) 向入口、 出口 节点发送故障通知；
3. 恢复信令传递时间， 包括发送时间和线路传播延时， 对于 GMPLS 扩展的 RSVP-TE信令协议， 恢复信令通过路径（Path ) 消息传递， 通 常每个消息 500 个字节左右， Path 消息是每通道实例的， 本发明描述的 恢复信令只通道一个字节或更小；
4. 协议处理时间， 从恢复信令接收到， 完成解码， 更新状态机， 转发 恢复信令， 这一过程需要任务间的通讯和其他的工作， 需要占有处理器时 间； 对应 GMPLS 扩展的 RSVP-TE信令协议处理相对复杂， 本发明描 述的恢复信令处理则非常筒单；
5. 交叉连接配置时间， 网元接收到恢复信令后需要配置交叉连接， 交 叉连接的调整动作需要时间 , 本发明中采用先转发恢复信令后启动交叉连 接的方法， 使得恢复路径上各网元的交叉连接动作并行化
本发明在因素 3 和 4 方面能节省处理时间， 加速恢复过程。 如果 GMPLS 扩展的 RSVP-TE 运行在 SDH 的开销字节 D4-12 组成的 576kbits/s 的 DCN上， 则传送 STM-64的 64 个 VC-4通道的恢复信令 的路径消息需要的时间为 500 x 8 x 64/576000 , 约为 440 毫秒。 本发明采 用的恢复信令发送方式， 如果直接采用开销字节传递， 则理论上讲在 125 微秒内完成所有 64 个通道的信令传递，加上网元的检测处理也应在 1 ms 左右完成； 如果采用上述的 DCN作为专用通道， 考虑到 HDLC分组头 的开销 64 个通道占有 60 个字节的恢复信令， 则发送时间在 50 X 8/576000, 约为 1毫秒。 在因素 4方面， GMPLS扩展的 RSVP-TE 的协 议处理开销相当复杂， 在数毫秒的时间， 而本发明描述的协议处理过程只 需要简单的查表过程， 处理时间可以忽略， 而 RSVP-TE在恢复信令按通 道串行处理则更加大了总开销。 另外， GMPLS使用的 IP DCN属于共享 性质的， 故障发生时需要进行的各种通知消息和路由协议的链路状态刷新 加重了 DCN 的负担， 使得基于 GMPLS 信令的恢复过程在时间上存在 4艮大的波动， 而本发明描述的通过专用恢复信令通道则可以避免这方面的 问题， 实现恢复时间的稳定和预测性。
虽然本发明是通过优选实施例进行说明的， 但是本领域技术人员应当 理解本发明并不限于这些实施例， 而是可以在不脱离本发明实质的情况下 对其进行各种变化和修改。 因此， 本发明的范围只由权利要求及其等同物 来确定。

Claims

权 利 要 求
1. 一种实现网格状光网络业务恢复的方法， 包括：
a) 为恢复路径在其经过的链路上预留通道资源并为各通道分配区分 器， 以及确定传送所述区分器的关联信令信道；
b)当检测出所述工作路径为故障时， 根据通过所述关联信令通道传送 的值为指示恢复的区分器激活所述恢复路径。
2. 根据权利要求 1所述的方法，其特征在于，按照下述步骤激活所述 恢复路径：
1 )恢复路径的入口网元查找恢复路径的下一跳网元、 利用所述关联 信令通道向下一跳网元发送值为指示恢复的区分器， 然后建立交叉连接； 2 )如果所述下一跳网元是恢复路径的中间网元， 则才艮据所接收的区分 器值查找再下一跳网元、 向所述再下一跳网元发送值为指示恢复的区分 器， 然后建立交叉连接， 直至出口网元收到值为指示恢复的区分器， 进入 步骤 3 ), 如果所述下一跳网元为出口网元， 则直接进入步骤 3 );
3 )所述出口网元建立交叉连接。
3.根据权利要求 1所述的方法， 其特征在于， 按照下述步骤激活所述 恢复路径：
1 )恢复路径的入口网元查找恢复路径的下一跳网元、 建立交叉连接， 然后利用所述关联信令通道向下一跳网元发送值为指示恢复的区分器； 2 )如果所述下一跳网元是恢复路径的中间网元， 则根据所接收的区分 器值查找再下一跳网元、 建立交叉连接， 然后向所述再下一跳网元发送值 为指示恢复的区分器， 直至出口网元收到值为指示恢复的区分器， 进入步 驟 3 )， 如果所述下一跳网元为出口网元， 则直接进入步骤 3 );
3 ) 所述出口网元建立交叉连接。
4.根据权利要求 2或 3所述的方法， 其特征在于， 进一步包括所述出口 网元沿所述恢复路径的相反方向发送确认恢复路径完成的信息的步骤。
5.根据权利要求 4所述的方法， 其特征在于， 所述信息为所述指示恢 复的区分器的值。
6.根据权利要求 2或 3所述的方法，其特征在于， 所述建立交叉连接 是指入口网元在连接的正向进行桥接， 在连接的反向将交叉切换到恢复路 径； 中间网元建立双向交叉连接； 出口网元在连接的正向将切换交叉到恢 复路径， 在连接的反向建立桥接。
7.根据权利要求 2或 3所述的方法，其特征在于， 所述网元通过恢复映 射表查找所述下一网元。
8.根据权利要求 7所述的方法， 其特征在于， 所述恢复映射表由所述 光网络的网管系统或规划系统计算或通过信令由各网元计算； 并由网络维 护人员逐网元设置或在为所述恢复路径分配区分器值时通过信令过程设 置。
9.根据权利要求 1、 2或 3所述的方法， 其特征在于， 所述关联信令通 道为所述光网络的物理帧结构中通道级开销的未使用的字节。
10.根据权利要求 1、 2或 3所述的方法， 其特征在于， 所述关联信令通 道为所述光网络的物理帧结构中服务层的开销中未使用的开销， 所传送的 区分器在发送物理帧的网元端口内唯一。
11.根据权利要求 1、 2或 3所述的方法， 其特征在于， 所述关联信令通 道为所述光网络的物理帧结构中服务层的开销中未使用的开销， 所传送的 区分器与所述恢复路径的通道间建立一一对应的关系。
12.根据权利要求 1、 2或 3所述的方法， 其特征在于， 所述关联信令通 道为运行在多个开销字节上的 HDLC协议信令通道。
13.根据权利要求 1、 2或 3所述的方法， 其特征在于， 所述区分器采用 比特位压缩编码。
14.根据权利要求 1、 2或 3所述的方法， 其特征在于， 通过服务层告警 检测来检测所述工作路径的工作状态， 或通过通道层告警来检测工作路径 的工作状态。
15.根据权利要求 14所述的方法，其特征在于， 所述服务层告警包括
SDH/SONET光网络的信令丢失、 帧丢失、 复用段告警指示、 复用段远端 告警指示、 复用段信号劣化告警、 及 OTN光网络的信号丢失、 远端告警 指示中的至少一种告警； 所述通道告警包括 SDH/SONET光网络的通道 / 虚路径的告警指示、 通道误码越限、 远端缺陷指示和远端故障指示告警、 及 OTN光网络的信号丟失、信号劣化和远端告警指示中的至少一种告警。
16.根据权利要求 1、 2或 3所述的方法， 其特征在于， 进一步在所述 工作路径修复后， 释放恢复路径上的交叉连接的步骤， 所述释放恢复路径 上的交叉连接的步驟包括：
1 )入口网元将业务将接收侧切换到工作路径， 沿恢复路径向中间网 元发送空闲区分器；
2 ) 中间网元接收到空闲区分器后， 根据恢复路径原来的区分器值查 找出通道， 向该通道发送的空闲区分器； 重复步驟 2 )直到出口网元收到 空闲态区分器；
3 ) 出口网元完成双向连接的正向的交叉连接切换， 实现双向业务两 个方向都切换到工作路径上， 拆除双向连接反向的桥接；
4 ) 所述出口网元沿恢复路径反向传递空闲态开销， 中间网元拆除两 个方向上的交叉连接， 入口网元拆除发送侧的桥接。
PCT/CN2005/001445 2004-09-10 2005-09-09 A method for optical meshwork service restoration Ceased WO2006026930A1 (en)

Priority Applications (2)

Application Number Priority Date Filing Date Title
EP05783726.2A EP1737145B1 (en) 2004-09-10 2005-09-09 Service restoration method for optical mesh networks
US11/568,766 US7965938B2 (en) 2004-09-10 2005-09-09 Method for mesh optical network service restoration

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CNB2004100737870A CN100555922C (zh) 2004-09-10 2004-09-10 一种实现网格状光网络业务恢复的方法
CN200410073787.0 2004-09-10

Publications (1)

Publication Number Publication Date
WO2006026930A1 true WO2006026930A1 (en) 2006-03-16

Family

ID=36036080

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2005/001445 Ceased WO2006026930A1 (en) 2004-09-10 2005-09-09 A method for optical meshwork service restoration

Country Status (4)

Country Link
US (1) US7965938B2 (zh)
EP (1) EP1737145B1 (zh)
CN (1) CN100555922C (zh)
WO (1) WO2006026930A1 (zh)

Families Citing this family (29)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN100555922C (zh) 2004-09-10 2009-10-28 华为技术有限公司 一种实现网格状光网络业务恢复的方法
CN1866806B (zh) 2005-12-22 2011-11-02 华为技术有限公司 共享格状网恢复的实现方法
CN100452739C (zh) * 2006-05-30 2009-01-14 北京邮电大学 一种面向网格应用的光并行传输方法
US20100238813A1 (en) 2006-06-29 2010-09-23 Nortel Networks Limited Q-in-Q Ethernet rings
CN101163030A (zh) * 2006-10-09 2008-04-16 华为技术有限公司 一种建立区分器映射表的方法
FR2909503B1 (fr) * 2006-12-04 2009-10-09 Alcatel Sa Procede d'etablissement d'une connexion bidirectionnelle
CA2673206A1 (en) * 2006-12-28 2008-07-10 Telefonaktiebolaget Lm Ericsson (Publ) Protection scheme
US8442030B2 (en) * 2007-03-01 2013-05-14 Extreme Networks, Inc. Software control plane for switches and routers
US8666242B2 (en) * 2007-06-05 2014-03-04 Cisco Technology, Inc. Response to OTUk-BDI for OTN interfaces to restore bidirectional communications
CN101399704B (zh) * 2008-11-05 2011-05-11 中国移动通信集团江苏有限公司 发现未开通上网功能但有需求用户的方法
US20100166420A1 (en) * 2008-12-22 2010-07-01 Electronics And Telecommunications Research Institute Apparatus and method for controlling route and resource in packet-optic convergence network
JP5267191B2 (ja) * 2009-02-18 2013-08-21 富士通株式会社 光リングネットワークシステム及び光伝送装置
CN101515827B (zh) * 2009-03-31 2011-12-07 中兴通讯股份有限公司 一种自动交换光网络业务错联阻错的方法及系统
CN101547382B (zh) * 2009-05-06 2012-01-04 烽火通信科技股份有限公司 一种格状网中非同源同宿主备业务资源分离的方法
CN102104495B (zh) 2009-12-18 2014-07-09 华为技术有限公司 共享Mesh保护的实现方法和设备及光网络系统
CN102868544A (zh) * 2011-07-06 2013-01-09 中兴通讯股份有限公司 一种实现共享网格保护的方法及系统
CN102412994A (zh) * 2011-11-23 2012-04-11 福建星网锐捷网络有限公司 接收设备、发送设备、线路故障处理方法及系统
US8718039B2 (en) * 2011-12-22 2014-05-06 Tt Government Solutions, Inc. Signaling protocol for multi-domain optical networks
US8854955B2 (en) 2012-11-02 2014-10-07 Ciena Corporation Mesh restoration and bandwidth allocation systems and methods for shared risk connection groups
JP6229318B2 (ja) * 2013-06-05 2017-11-15 富士通株式会社 通信システム、通信制御方法、及び、伝送装置
ES2607496T3 (es) * 2014-02-25 2017-03-31 Alcatel Lucent Detección de fallos de GCC rápida
BR122023005931B1 (pt) * 2014-10-07 2024-01-02 Sedonasys Systems Ltd Sistemas para gerenciamento de uma falha e de uma rede de comunicação
CN108270648A (zh) * 2016-12-30 2018-07-10 中国移动通信集团内蒙古有限公司 城域骨干传送网otn组网方法
US10536216B1 (en) * 2018-07-24 2020-01-14 Ciena Corporation Service synchronization in retain home path scenarios in a control plane network
US11178218B2 (en) * 2019-01-02 2021-11-16 Citrix Systems, Inc. Bidirectional communication clusters
US10862795B2 (en) * 2019-01-08 2020-12-08 Fujitsu Limited Provisioning recovery paths in a mesh network
US11283518B2 (en) * 2019-11-08 2022-03-22 Infinera Corporation Method and apparatus for a restoration network with dynamic activation of pre-deployed network resources
CN113810793A (zh) * 2020-06-12 2021-12-17 中兴通讯股份有限公司 资源部署方法、设备、管控系统及计算机存储介质
WO2024030135A1 (en) * 2022-08-05 2024-02-08 Altiostar Networks, Inc. System and method for traffic distribution with optical switch

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP1411666A2 (en) * 2002-10-16 2004-04-21 Samsung Electronics Co., Ltd. Two-fiber optical ring network
CN1499747A (zh) * 2002-11-07 2004-05-26 华为技术有限公司 一种实现智能光网络保护和恢复的方法

Family Cites Families (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5986783A (en) * 1997-02-10 1999-11-16 Optical Networks, Inc. Method and apparatus for operation, protection, and restoration of heterogeneous optical communication networks
KR100537746B1 (ko) * 2001-01-10 2005-12-19 학교법인 한국정보통신학원 광인터넷에서의 ＭＰλＳ 보호 및 절체방법
US7103044B1 (en) * 2001-04-17 2006-09-05 Keller Richard B Utilizing available SONET overhead bytes for additional signaling channels
US6952395B1 (en) * 2001-06-01 2005-10-04 Calix Networks, Inc. Optical network restoration
US7652983B1 (en) * 2001-06-25 2010-01-26 At&T Intellectual Property Ii, L.P. Method for restoration and normalization in a mesh network
KR100462408B1 (ko) * 2002-12-10 2004-12-17 한국전자통신연구원 Ｇｍｐｌｓ를 통한 빠른 재 루트 방법
CN100555922C (zh) 2004-09-10 2009-10-28 华为技术有限公司 一种实现网格状光网络业务恢复的方法

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP1411666A2 (en) * 2002-10-16 2004-04-21 Samsung Electronics Co., Ltd. Two-fiber optical ring network
CN1499747A (zh) * 2002-11-07 2004-05-26 华为技术有限公司 一种实现智能光网络保护和恢复的方法

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
"Internet Draft, Work in progress", May 2004, article "RSVP-TE Extensions in support of End-to-End GMPLS-based Recovery"
See also references of EP1737145A4 *

Also Published As

Publication number Publication date
EP1737145A4 (en) 2007-09-05
US7965938B2 (en) 2011-06-21
EP1737145B1 (en) 2014-06-18
EP1737145A1 (en) 2006-12-27
CN100555922C (zh) 2009-10-28
US20090310960A1 (en) 2009-12-17
CN1747368A (zh) 2006-03-15

Similar Documents

Publication Publication Date Title
WO2006026930A1 (en) A method for optical meshwork service restoration
US7787362B2 (en) Method and device for recovering a shared mesh network
CN101447849B (zh) 网络恢复
JP5163479B2 (ja) パス切替え方法
EP1845656B1 (en) A method for implementing master and backup transmission path
EP2302836B1 (en) Method, system and device for recovering service
EP1735950B1 (en) Line-level path protection in the optical layer
US20080056159A1 (en) Method for setting path and node apparatus
JP2003298633A (ja) 制御チャネル障害時のデータチャネル障害通知機能を有する伝送装置
CN1833382A (zh) 光网络网间互通结构及其连接配置方法
WO2011026442A1 (zh) 光网络中的信息处理方法、光通信装置及系统
WO2008119294A1 (en) Method and equipment for restoring network business
CN101176303A (zh) 一种业务倒换的方法和网络节点
WO2011017941A1 (zh) 网络保护方法及装置
WO2006063512A1 (fr) Procede de realisation de l'operation de commutation d'un reseau en anneau a origines et destinations multiples
WO2007085173A1 (en) A method for processing network resource, a network unit in an intelligent optical network thereof
CN1499747A (zh) 一种实现智能光网络保护和恢复的方法
CN101001123B (zh) 用于光传输网络的倒换方法和装置
US20030043427A1 (en) Method of fast circuit recovery using local restoration
CN100531092C (zh) 智能光网络的业务重路由触发方法
WO2011017863A1 (zh) 光数据单元环路共享保护控制方法与装置
CN100382534C (zh) 智能光网络双向复用段环网络保护倒换失败的检测方法
EP2219328B1 (en) A method for soft re-routing in ASON
WO2008040254A1 (en) The process method for traffic engineering link information
CN1710993B (zh) 智能光网络中的业务路径优化方法

Legal Events

Date Code Title Description
AK Designated states

Kind code of ref document: A1

Designated state(s): AE AG AL AM AT AU AZ BA BB BG BR BW BY BZ CA CH CN CO CR CU CZ DE DK DM DZ EC EE EG ES FI GB GD GE GH GM HR HU ID IL IN IS JP KE KG KM KP KR KZ LC LK LR LS LT LU LV MA MD MG MK MN MW MX MZ NA NG NI NO NZ OM PG PH PL PT RO RU SC SD SE SG SK SL SM SY TJ TM TN TR TT TZ UA UG US UZ VC VN YU ZA ZM ZW

AL Designated countries for regional patents

Kind code of ref document: A1

Designated state(s): BW GH GM KE LS MW MZ NA SD SL SZ TZ UG ZM ZW AM AZ BY KG KZ MD RU TJ TM AT BE BG CH CY CZ DE DK EE ES FI FR GB GR HU IE IS IT LT LU LV MC NL PL PT RO SE SI SK TR BF BJ CF CG CI CM GA GN GQ GW ML MR NE SN TD TG

121 Ep: the epo has been informed by wipo that ep was designated in this application
WWE Wipo information: entry into national phase

Ref document number: 11568766

Country of ref document: US

Ref document number: 2005783726

Country of ref document: EP

WWP Wipo information: published in national office

Ref document number: 2005783726

Country of ref document: EP

NENP Non-entry into the national phase

Ref country code: DE