There is no question that fiber optic communication is our future. Fiber optic communication industry has been enjoying amazing growth for over 15 years. These are driven by both technology advance and market demand. There are some obvious trends in the development of new technology and market. Let's examine some of the most important ones here.
All-optical network has been a top topic in fiber optic communication industry for over a decade now. Its ultimate goal is to process all signals in the optical domain without any conversion and controlling to electrical domain at all.
At least for now, most signal routing, processing and switching happens in the electrical domain. Optical signals have to be converted to electrical signal first, and then the electrical signals are processed, routed and switched to their final destination.
After the processing, routing and switching, the electrical signals are then converted back to optical signals which are then transmitted over long distances. This process is called the O-E-O process.
But this O-E-O process severely limits the speed of the network. Why? Since optical signals can process data much faster then today's electronics. The O-E-O process has been a bottleneck preventing us from achieving even higher data rates.
This bottleneck creates a tremendous interest in all-optical networks where no electronics are needed for signal processing, routing and switching. Another big benefit of all-optical network is that since all signal processing, routing and switching happens in optical domain, there is no need to replace the electronics when data rates increase. For example, current fiber optic transmitters and receivers can handle only one single data rate, thus, they must be replaced when the data rate increases. This won't be necessary in a all-optical network.
However, many obstacles still lie in our way to make all-optical network a reality. Some functions such as reading headers on the optical signals, switching the optical signal on the fly based on the header content and real-time wavelength switching are just a few of the serious challenges that need to be solved before we can have a true all-optical network.
DWDM opens the door to multi-terabit transmission. The interest in developing multi-terabit networks is driven by the increasing availability of more bandwidth in fiber optic networks.
One terabit network was achieved by using 10Gb/s data rate combined with 100 DWDM channels.
Four terabit network can be achieved by combing 40Gb/s data rate with 100 DWDM channels too. Researchers move their target to even higher bandwidth with 100Gb/s systems. (Which is not a reality yet, at least for now)
But this kind of speed is very expensive to make and can only be justified on long-haul systems. But with the cost reduction on fiber optic components, devices and systems, more bandwidth is not far from us.
There are some other major trends in the fiber optic industry too. The most important ones include expansion into mass markets (FTTH, FTTB, FTTC, etc), miniaturization, new technology development, cost reductions and even more.