After last week writing about the upgrade paths for PON networks, this week’s post will be a technical one about WDM PON technologies.
WDM PONs are a good way to upgrade existing PON topologies, by adding more optical channels to the access segment, thus increasing overall bandwidth capabilities.
The main principle of WDM PON is to establish a logical point-to-point link from the Central Office to each subscriber, by using a dedicated wavelength to every CPE. It seems easy, but there are several technological challenges to overcome.
The first advantage of WDM PONs are their transmission capabilities. As each subscriber has a dedicated wavelength, the potential bandwidth per user dramatically increases. One can allocate a Gigabit or a 10 Gigabit connection to each CPE quite easily. At the end of the day, it is just a matter of using the right optical interfaces. Also, network flow management is simplified as the infrastructure logically behaves as a very simple point to point network.
However, it is not everything nice and simple. The present major drawbacks of WDM PONs at present are related to cost of the required components.
At the remote node, a wavelength multiplexer to separate the different wavelengths to each subscriber is required and those are not cheap devices at present. Although they have improved a lot during the last years and now they do not have thermal stability issues as in the past, their cost is still higher than traditional power splitters.
Talking about CPEs, cost is also higher for WDM PON devices. Their technology is more advanced than in TDM PON gear and additionally, their economies of scale are smaller at present. They can potentially become quite similar in price with TDM PON boxes, but there is a technical challenge to overcome: how to transmit at an specific wavelength (each CPE uses a different wavelength to transmit to the CO) without requiring a tunable laser (because its price is simply too high). This challenge has already been solved by using mode-locked transmitters but there is still room to reduce costs.
Other considerations to be taken are environmental and related to power dissipation at the CO. The OLT is a technologically equivalent to a point-to-point one and therefore the number of optical interfaces is quite high (one per subscriber). This makes that the power requirements at this end are higher than in TDM PON technologies (2W per user against 0.5W of TDM PON systems).
In any case, the standardization bodies have started to work in order to define WDM PON standards. ITU is already working on an evolution of ITU-T G.987 10G PON standard in order to support WDM PON architectures. Additionally, at present, several manufacturers offer their WDM PON solutions, but in most of the cases one needs to buy the end-to-end system to make it work.
I believe that WDM PON is a candidate to be deployed in FTTH networks. Not now, because bandwidth demands do not justify it unless in a corporate environment but in the mid term, once the price of WDM PON components are viable for massive deployments.