What is the OSA-710?
It is an instrument that allows in-service support of 100G channels in your terrestrial backbone network without needing to turn off services to measure the optical signal-to-noise ratio (OSNR). This solves the need to verify the optical performance of the network before proceeding with BER testing execution.
Why measure in-service during the life of a 100G service?
It happens that to deploy and maintain your optical network of 100G services, the traditional OSA instrument can only perform OSNR measurement through the on/off method. While this is not a problem during the network construction and commissioning phase, it will be, and very much so, when maintenance is required during the service stage. Even during the night, in a maintenance window, turning off a 100G channel means cutting service to many customers…it is no longer an easily implementable option.
Will it allow me to save costs?
It is precisely to eliminate costs that this new OSA is recommended. It serves equally during deployment and commissioning, as well as for later assignment to your network operations. At this stage, it will guarantee visibility and control of your traffic. The guarantee lies in its method, which makes it independent of the service type.
Measure the OSNR of any service, on any channel, without ever turning anything off..??
Exactly, the OSA-710 uses the so-called "Spectral Correlation method" which is based on having a bandwidth with ultra-high resolution (200 times better than any other OSA) that ensures the ability to measure Nyquist signals and super-channels. It is also independent of the line rate (100G, 200G, 400G), the modulation scheme, or the use of POL-MUX techniques. Regarding the latter, it can be seen that the OSA solves the problem of measuring OSNR when the format is polarized, as it does not require the use of null-polarization techniques (which were used by previous OSAs)
Is there another way to make these measurements?
Certainly some ideas about other OSA methods have emerged to be able to measure these new services. For example, if one could assume that the TX laser is a clean point of reference without noise, and that it provides the correct information of the spectral waveform of the signal…then one would measure and save that curve, then at any point where one wants to measure, would subtract it from the measured signal and would obtain spectral information of the noise, accessing OSNR indirectly…. But is this a good assumption? Unfortunately not, because the ROADM link includes optical filters, the reference information becomes decorrelated inevitably! It would be necessary to know the OSNR at the output of each ROADM to be able to maintain a valid reference of the original signal. Since those OSNRs are precisely part of what we want to measure, this appears to be circular reasoning.