CN100421397C - A performance and fault monitoring method of an optical burst switching network - Google Patents

Abstract

The invention claims protection of an optical burst switching network performance and fault monitoring method, which relates to the technical field of optical communication. The method generates a detection burst PB and a corresponding detection burst control packet PBHP on the data channel of the optical burst switching network, and reflects the fault of the optical layer to the electrical layer through the detection burst, and monitors the performance index of the network online. Under the control of the control plane or management plane, periodically send PB on the data channel to be detected, and send its corresponding PBHP on the control channel in advance, receive and take out the PB at the next-hop network node of the sending PB node, and Check the bits in the PB, and statistically analyze the bit errors. If the fault alarm threshold is exceeded, the network element will send an alarm message. The invention can quickly reflect the fault of the optical layer to the electric layer by detecting the burst, and can provide fast and accurate fault information for the fast protection switching or recovery required by the network survivability and fault location.

Description

A performance and fault monitoring method of an optical burst switching network

technical field

The invention relates to the technical field of communication, in particular to the fault monitoring of an optical interconnection network.

Background technique

Optical burst switching technology combines the technical advantages of optical circuit switching and optical packet switching, and is one of the preferred solutions for the next generation of all-optical Internet. The optical burst switching network can be regarded as a superimposed network of the electric domain control network and the all-optical data transmission network. It uses electrical signals to control the optical cross-connect (OXC) equipment to complete the fast cross-connection of the wavelength channel, which is considered to be very A promising form of wavelength division multiplexing (WDM) network switching. It has always been a hot research direction to improve the monitoring mechanism of the optical domain so that the monitoring parameters of the optical domain can ensure the normal operation of this layer. Such as literature: Wen, Y.; Chan, V.W.S.; Efficient Fault-Diagnosis Algorithms for All-Optical WDM Networks With Probabilistic Link Failures, Zheng, L. Lightwave Technology, IEEEJournal of Volume 23, Issue 10, Oct.2005, but due to the current Optical logic functions are very limited, optical buffer technology is not yet mature, and the improvement of optical domain monitoring parameters will also be a long-term process. Before the monitoring mechanism of the optical domain is mature, the optical domain must complete its own survivability monitoring such as bit error monitoring, connectivity status, and remote defect indication by the carried electrical domain services. For example, the SDH system in the SDH over WDM system completes these functions , When ATM is transmitted on WDM, ATM completes this function, etc. In these systems, the optical signal is subjected to a single-hop test (single-hop test) at each node, that is, the optical signal is converted into an electrical signal for bit verification, signal regeneration, etc., and then sent out through electro-optical conversion. faults can be detected within one hop. However, because the optical burst switching network data channel transmits a transparent all-optical signal. Moreover, the electric domain service cannot guarantee the quality of the carried signal. It is difficult to implement the method of monitoring channel performance by verifying signals at the electrical layer in an all-optical transparent network. Therefore, the above-mentioned existing monitoring method is not applicable to the optical burst switching network.

Contents of the invention

The technical problem to be solved by the present invention is: Aiming at the current situation that the data channel in the optical burst switching network transmits transparent all-optical signals and there is no perfect fault detection means in the prior art, the present invention proposes a new optical burst The performance monitoring mechanism of the data channel of the optical burst switching network. The mechanism sends the detection burst to the data channel periodically or in a specific way at the network element node of the optical burst switching network, and monitors the detection burst at each core node hop by hop. To evaluate the status of the data channel between each two nodes, to judge the soft and hard faults of the monitored network.

The technical solution adopted by the present invention to solve the technical problem is: to artificially generate the detection burst signal PB (Probe Burst) and corresponding The PBHP (Probe Burst Header Packet), which detects burst signals, can quickly reflect faults at the optical layer to the electrical layer by detecting burst signals, thereby monitoring network performance indicators online. The present invention can further monitor the soft and hard faults of the optical burst switching network, and the fault monitoring method specifically includes the following steps: generating a probe burst signal PB (Probe Burst) and a corresponding detection burst signal at each network element node of the optical burst switching network The control packet PBHP (Probe Burst Header Packet) of the burst signal; under the control of the control plane or the management plane, periodically (or in other prescribed ways) send PB with a fixed bit pattern detection frame on the data channel to be detected, And send its corresponding control packet PBHP on the control channel in advance; take out the PB at the next hop network node of the sending PB node, and check the bits in the PB, statistically analyze its error status, and store it in the error status log; If the code error deteriorates extremely and exceeds the hard fault alarm threshold, the network element will send a hard fault alarm to the protection recovery module; compare and analyze the code error situation with the code error history, if the code error situation exceeds the soft fault alarm threshold, then the network element will Send a soft fault alarm to the network management center.

The beneficial effect of the present invention is that it does not require complex external monitoring equipment, and it can provide fast and accurate alarm information for fast protection switching or recovery required for network survivability; it can also provide accurate fault location for management and maintenance purposes. Detailed soft and hard fault information, and it is applicable to any network scale, any network topology and control mode, and has good scalability.

Description of drawings

Figure 1 Data burst DB, detection burst PB signal format example

Figure 2 (a) Functional diagram of the entry part of the edge node (b) Functional diagram of the exit part of the edge node

Figure 3 Functional structure diagram of the core node with detection module

Figure 4 Schematic diagram of the transmission of PB and DB in OBS

Figure 5 Flowchart of Fault Monitoring Principle

Detailed ways

Hard faults in optical networks are often caused by fiber breaks, power failures, damage to lasers or receivers, and human attacks. . Soft faults are often caused by device aging, temperature drift, etc., and are manifested as an increase in the number of random bit errors per unit time (random bit error rate). The state of the channel can be estimated by sampling the channel. In the data channel, sending detection bursts periodically or in a specific way is equivalent to sampling the state of the data channel to detect burst errors. code condition to assess the state of the data channel. If the detection frame encapsulated in the detection burst has a fixed bit pattern, then after receiving the PB, the receiving end only needs to compare the detection frame encapsulated in the PB with the standard pattern to know that the error in the PB is a random error. Or a sudden error. As shown in Fig. 4, in addition to transmitting the normal service DB, the data channel group DCG is also responsible for transmitting the detection burst PB between adjacent network elements. The control packets DBHP (and PBHP) of the DB (and PB) are transmitted on the control channel group CCG. DB enters the data channel group from the entrance edge node, goes straight to the core switching node, and ends at the exit edge node, without any photoelectric conversion in the middle. PB is different. PB can enter the data channel group from any network element node, but it must be dropped from the next hop network element node and sent to the Probe Process Module for verification.

1. Format of data burst and detection burst

Figure 1 shows the data burst DB of the optical burst switching network and the corresponding data burst control information packet DBHP (Data Burst Header Packet), the detection burst PB (Probe Burst) and the corresponding detection burst control packet PBHP (Probe Burst) Burst Header Packet) data structure diagram. Figure 1(a) shows the signal format of DB and its DBHP. After the upper-layer business data is aggregated and processed in the burst adaptation layer, a data burst DB (Data Burst) and a corresponding data burst control packet DBHP (Data Burst) are generated. Burst Head Packet). The information contained in DBHP generally includes the source address (Src) of the DB, the destination address (Des), the wavelength number (Wavelength) to be used, the size of the DB (Size), and the offset time (Offset time). Sent to the physical layer of the control network for transmission. The DB is sent to the data network as a payload, and it is generally not read or rewritten by the control part of the node, so it is often called "data transparently passes through the network". Different from DB, the payload of the probe burst is not generated by aggregation. It is generated by the probe module (Probe Module) based on the principle of fault monitoring and encoding, and has a strong ability to monitor burst errors. PB The signal format of PBHP and its PBHP is shown in Figure 1(b). It adopts the same burst format as DB, including the same control packet format and payload format. In the payload part, what is filled is not the upper layer business, but a Detection frame, this frame has a strong ability to monitor burst errors, and the node distinguishes PB and DB through the type (type) byte in the control packet.

2. The processing flow of burst packets

As shown in Figure 2(a), at the sending part of the edge node (edge entry node). The assembly module (Assembler) completes the aggregation of business data to form a DB; the probe module (Probe Module) generates PB according to the established rules. When the DB or PB is generated, the BHP generator (BHP Generator) generates the corresponding burst control packet. The burst control packet of DB is DBHP, and the burst control packet of PB is PBHP. At this time, the control packet in the electrical signal format undergoes E/O conversion, becomes an optical signal, and is sent to the control channel group. After the offset time arrives, the Offset Manager controls the scheduler to send bursts. At this time, the burst packet in the electrical signal format is converted into an optical signal by E/O and sent to the data channel group. As shown in Figure 2(b), at the receiving part of the edge node (edge egress node), the signal on the optical fiber is converted into an electrical signal by the receiver (O/E conversion). The BHP Processor (BHP Processor) controls the digital cross-connect device according to the burst type information provided by the BHP, so that the DB is sent to the disassembler (disassembler), and the PB is sent to the detection processing module.

Figure 3 is a functional structure diagram of a core node with a detection module. At the core switching node, the switching control unit SCU receives the BHP on the control channel, and configures the optical cross-connect device OXC according to the address and wavelength information in the BHP, so that the corresponding DB on the data channel is switched to the correct egress wavelength, or dropped sent to the edge exit node. If the SCU processes the PBHP, it controls the OXC to switch the PB to the detection module, and the detection module processes the detection burst and evaluates the performance of the data channel. According to the requirements of the management or control plane, the SCU can also generate the PBHP of the egress wavelength channel to be detected and notify the detection module to prepare PBs. After a suitable offset time, the SCU controls the OXC to schedule the corresponding PBs to be sent to the corresponding egress wavelengths. The dotted line in FIG. 3 indicates that the control channel is one wavelength or several wavelengths among all wavelength channels (control channel group CCG). Figure 4 can be used to represent the decoupling relationship between the data channel group (DCG) and the control channel group in the OBS. It is worth reminding that PB is generally relatively short, and its survival period is only one hop; while DB is generally relatively long, and its survival period generally has multiple hops from the entry node to the exit node.

Both edge and core nodes of the optical burst switching network can complete the generation, transmission, reception and judgment of network performance of detection bursts.

3. Fault monitoring process

The detection frame encapsulated in the detection burst can use a fixed bit pattern. After receiving the PB, the receiving end only needs to compare the detection frame encapsulated in the PB with the standard pattern to know whether the error in the PB is a random error or an error. sudden error. As shown in Figure 5. Take out the PB at the next-hop network node of the sending PB node, check the bits in the PB, statistically analyze the error status, and store the error status in the statistical log. If the received frame format is incorrect, or the detection frame When there is a sudden error or when PBHP should receive PB but not receive it, it will cause the up-down counter to increase by one, but if the correct frame is received next, the up-down counter will be decremented by one. When the count value increases and exceeds the previous When the hard fault alarm threshold is set, it means that the bit error is extremely deteriorated, and the network element will send a hard fault alarm to the protection and recovery module; count the number of errored frames caused by random errors, the number of errored frames per unit time, and compare the number of errored frames with the previous , Frame error rate and other historical records are compared and analyzed. If the bit error exceeds the soft fault alarm threshold, the network element will send a soft fault alarm to the network management center. A rising random bit error rate often indicates aging of the device.

The invention can quickly reflect the failure of the optical layer to the electrical layer by detecting the burst. It can provide fast and accurate alarm information for fast protection switching or restoration required for network survivability; it can also provide accurate and detailed soft and hard fault information for fault location for management and maintenance purposes. And it is applicable to any network scale, any network topology and control mode, and has good scalability.

Claims (4)

1. A kind of performance of optical burst switching network and fault monitoring method, it is characterized in that, at each network element node of optical burst switching network, produce the control packet PBHP of detection burst signal PB and detection burst signal; Under the control of the control plane or the management plane, periodically send the PB with a fixed bit pattern detection frame on the data channel to be detected, and send the corresponding PBHP on the control channel in advance; receive it at the next hop network element node sending the PB And take out the PB; check the bit pattern in the PB, and statistically analyze its error status; if the fault alarm threshold is exceeded, the network element will send an alarm message. 2. The performance and fault monitoring method according to claim 1, wherein the generation, transmission and reception of the detection burst signal are all completed at the edge and core nodes of the optical burst switching network. 3. The performance and fault monitoring method according to claim 1, wherein the lifetime of the detection burst signal is only one hop. 4. According to the performance and fault monitoring method described in one of claims 1-3, it is characterized in that, if the bit error situation exceeds the hard fault alarm threshold, the hard fault alarm information is sent to the protection recovery module; if the bit error situation exceeds the soft Fault alarm threshold, send soft fault alarm information to the network management center.

Images