The chain network of a customer consists of 19 stations (S1�S19). S1 and S19 are OTM end stations, S8 and S14 are pass-through (balancing) stations from the M40v to the D40, and other stations are OLA stations. Its networking is shown in an attachment (Figure 1.jpg).
There are 50 waves of 10 Gbit/s services from S1 to S19. After the power is adjusted, the error rate before correction of certain wavelengths reaches 10e-5. A hidden trouble exists during service running.
The investigation result shows that online engineers use the traditional receive-end adjustment method to adjust the receive optical power of S8, S14, and S19 when balancing the power, and the method is not optimal. The intermediate station adjustment method is recommended.
Currently, three methods are available to enable signals to keep optical flatness during transmission.
1. The first method is the intermediate station adjustment method. As shown in Figure 2, perform monitoring at OLA station B between two stations that have the power balancing function, and adjust the M40v at the transmit end to adjust the power of OLA station B.
2. The second method is to input a pre-balancing curve at the transmit end. The pre-balancing curve is summed up through the test at a laboratory, and is the relative difference of each channel.
3. The third method is to adjust one half at the transmit end. As shown in Figure 2, perform monitoring at station C and perform adjustment at station A.
(1) Adjust the transmit power at station A of the transmit end, and adjust the total receive optical power at station C. Compare the received power of each channel with the nominal power, and then obtain one half of the difference Xi.
(2) Adjust the power Xi of the corresponding channel at the transmit end.
The first method requires that engineering personnel arrive at the intermediate OLA station or configure the MCA (currently, the configuration is unavailable) at the OLA station. With regard to stations that have multiple spans with different spacing, it is difficult to find the accurate intermediate station. A difference exists for different fibers and networking situations when the second method is used. Therefore, the adjustment result is not accurate. The third method does not require arrival at the intermediate OLA station, thus facilitating operations. When handling a problem on the existing network, Huawei engineers cannot arrive at the intermediate OLA station. Therefore, use the third adjustment method. The adjustment process is as follows:
With regard to the spans of S1�S8: Adjust the transmit power of S1, monitor the receive power at S8, compare each wavelength power with the nominal power, obtain one half of the difference Xi, and adjust the attenuation value Xi of the corresponding wavelength at the transmit end of S1.
With regard to the spans of S8�S14: Adjust the transmit power of S8, monitor the receive power at S14, compare each wavelength power with the nominal power, obtain one half of the difference Xi, and adjust the attenuation value Xi of the corresponding wavelength at the transmit end of S8.
With regard to the spans of S14�S19: Adjust the transmit power of S14, monitor the receive power at S19, compare each wavelength power with the nominal power, obtain one half of the difference Xi, and adjust the attenuation value Xi of the corresponding wavelength at the transmit end of S14.
During the transmission of optical signals, the intrinsic insertion loss of a fiber for optical signals at band C is that the loss of a longer wavelength is shorter than the loss of a shorter wavelength, thus causing the slope of a power spectrum shape. The fiber Raman effect increases the slope. The gain of the optical amplifier for each wavelength is not the same strictly. Therefore, the power of each wavelength is different after the system transmission. The over high power in the system will cause the non-linear effect. The over low power will cause that the OSNR is slightly low. Thus, power balancing is required to prevent the over large difference of the channel power.
The traditional method of commissioning WDM power balancing is to look at the receive end and adjust the transmit end to adjust the optical power at the receive end. This is a simple onsite power adjustment method, which is used on the network whose span number is smaller than six. The other method, namely, link intermediate power adjustment method, is to adjust the power between two stations (OTM, FOADM, ROADM, REG, or OEQ stations) that provide the power balancing function. When the link intermediate power adjustment method is used, the difference between the maximum power and the minimum power in the line is about one half of the difference by using the original method, and the impact of the non-linear factor is reduced, the noise base in a link is improved, and favorable OSNR flatness is obtained. For comparison of the impacts of two balancing methods on the optical power in the link, see an attachment (Figure 2.jpg).
Therefore, on the 40 Gbit/s transmission network and the 10 Gbit/s multi-span network, the link intermediate power adjustment method is used to balance the power.
After optimization, the error rate before correction of each wavelength of the system improves by one to two levels. The error rate before correction of the channel with the poorest performance also reaches 10e�7. Thus, the power is balanced.
1. During link power commissioning, note that the total optical power cannot exceed the nominal optical power regardless of the method you used.
2. On the basis of power balancing, the BER at the receive end should meet service running requirements. When the BER of certain wavelengths is comparatively poor and the power needs to be improved, use the power of the wavelength with the well BER for balancing.