40G Service Provisioning Fails On a Network, The OTU_LOF And DCM_INSUFF Alarms A

Procedure

1. Connect an optical spectrum ***yzer to the receive end. It is found that the OSNR at the receive end is higher than 18 dB, which complies with the design value and satisfies the transmission requirement for 40G boards. Therefore, the fault is not caused by a low OSNR. Theoretical OSNR and actual OSNR are given below.

   

2. Replace the fiber jumper connecting the LUR40S to the V40 at the transmit end and the fiber jumper connecting the LUR40S to the D40 at the receive end, and replace the LUR40S boards at the receive and transmit ends. In this case, services are still unavailable. Therefore, the fault is not caused by faults on the LUR40S boards or fiber jumpers.
3. Query and then optimize the optical power of the line and the LUR40S boards. In this case, services are still unavailable. Therefore, the fault is not caused by abnormal optical power.
4. Test fibers before provisioning 40G services. The data collected in the fiber test shows that both the reflection measured by using the OTDR and the DGD measured by using the PMD meter satisfy the requirement for transmitting 40G services. Therefore, the fault is not caused by abnormal PMD or reflection. In addition, the test by using the OTDR shows that the type and length of fibers comply with design values.
5. Based on the preceding analysis and tests, it is determined that the fault may be caused by incorrect dispersion. In the design of dispersion configuration, the length of the span from sites F04 to F25 is 665.1 km, but a 660 km DCM is used on the span. Therefore, 5.1 km is not compensated. This is a good dispersion compensation solution to 40G services. It is found that the DCMs are reconfigured twice on this span, and therefore the DCMs may be mis-placed during the reconfiguration. Then, to prove this suspect, dispersion compensation of a single wavelength is configured at each end for testing.
6. In an analysis, it is found that that 40G services are available in the direction from sites F25 to F04 after positive DCM A is configured to compensate for a single wavelength. (A positive DCM refers to a DCM with a positive dispersion coefficient, which is opposite from that of a common DCM. When the positive DCM is successfully applied, it indicates that the original link is over-compensated.) In this case, the TDCM built in the LUR40S successfully searches for a dispersion compensation value, and the optimal dispersion compensation value is 60 ps. Therefore, over-compensation occurs in this direction, and the over-compensation is calculated as follows: 20 + (60/16.5) = 23.63 km
7. In the direction from sites F04 to F25, it is found that that 40G services are available in the direction from sites F25 to F04 after DCM A is configured to compensate for a single wavelength. In this case, the TDCM built in the LUR40S successfully searches for a dispersion compensation value, and the optimal dispersion compensation value is –230 ps. Therefore, under-compensation occurs in this direction, and the under-compensation is calculated as follows: 20 - (230/16.5) = 34 km. When the optimal dispersion compensation of the TDCM is 60 ps, it indicates that the link is over-compensated. The typical dispersion coefficient of G.652 fibers is 16.5 ps/nm.km, and therefore over-compensation is 3.63 km (60/16.5 = 3.63 km). Likely, when the optimal dispersion compensation of the TDCM is –230 ps, it indicates that the link is under-compensated. Therefore, under-compensation is 14 km (230/16.5 = 14 km). Theoretical residual dispersion and actual residual dispersion are given as below.

    

8. According to the preceding fault identification, it is determined that the DCMs on the line are abnormal. As a result, in the direction from sites F25 to F04, the dispersion compensation distance (5.1-(-23.63)=28.64 km) is 30 km, which is longer than the theoretical value; in the direction from sites F04 to F25, the dispersion (5.1-34=-28.9 km) is 30 km shorter than the theoretical value.
9. In one direction, dispersion compensation distance is 30 km, which is longer than the theoretical value; in another direction, dispersion compensation distance is 30 km, which is shorter than the theoretical value. This indicates that the DCMs on the line are probably mis-placed. In the design drawing, it is found that the fault most probably lies in site F25. If DCMs B and T at site F25 are reversely placed, the fault occurs. According to the calculated insertion loss (input optical power of the VA4 minus input optical power of the OBU or input optical power at optical interface 3 on the OAU minus input optical power at optical interface 5 on the OAU) of the two modules, it is found that the insertion loss of module B is 1.4 dB and the insertion loss of module T is 2.4 dB. (Determine whether the two modules are reversely placed by comparing the insertion loss of them. If they are correctly inserted, the insertion loss of module B must be higher than that of module T). Therefore, the two modules are probably reversely placed. In addition, it is found that dispersion of this site was adjusted two years ago. Therefore, it is almost certain that the fault results from reverse placement of the two modules. Dispersion configuration at site F25 is given as followed.

   

10. Based on the preceding analysis, go to the site again and carefully check fiber connections. Finally, it is determined that modules B and T are reversely placed.