In the first installment on coherent optics, we learned about the benefits of coherent optics; very high spectral efficiency and large link budgets. Yet, coherent communication has not found a niche in consumer level communications, such as PON systems.
The reason for this may be intuitively clear when you consider the number of precision optical components needed for a conventional coherent receiver as shown in the figure below (other approaches to receiver design are possible, some relying heavily on powerful DSPs). In this receiver, there are 21 passive components, four photodetectors and one highly stable tunable local oscillator. When combined with the need for an equally stable laser transmitter, the cost of the transmitter and receiver components becomes too high for a consumer level product.
Coherent optics were briefly considered for NG-PON2 due to the large link budget and ability to space wavelengths much more closely than DWDM systems. Due to the complexity described above and lack of evidence that the optics could be made cost effective, the ITU standards body decided to select the TWDM-PON technology for NG-PON2.
In the next installment, we will examine a modified form of coherent optics called “Quasi-Coherent Optics” that might change the cost equation for PON systems, particularly NG-PON2.