Now, we’re seeing P-OTS move into the core of the network. What’s driving this new trend? The core transport network has largely been ignored for a couple of years while the industry has focused on rebuilding access, metro, and aggregation networks for packets and IP. The reality is that growth in Ethernet and IP, Internet data and video, and mobile data and video impact core networks as well as metro access networks. The migration from Sonet/SDH to packets is not just a metro problem.
The biggest operator transport challenge in the core right now is the heavy traffic burden being borne by core routers that process all of this IP traffic moving through the core. Operators tell us that as much as 60 percent of traffic hitting core routers is transit traffic, meaning that, for up to 60 percent of the traffic, the heavy IP processing performed by core routers is not necessary.
This traffic mismatch is significant because core router ports are tremendously expensive when compared to Ethernet and optical equipment port costs. For example, we’re told that router vendors are pricing 40G router ports at about 10 times the cost of 40G DWDM transponders. As a result, in core networks, core router bypass is becoming the new buzzword. Added incentives for core router bypass are the power and footprint savings to be realized – both increasingly important in the new era of “green telecom.” As a rule, optics consumes less power and occupies less space than active electronics.
Verizon has done an excellent job defining the requirements for these new core packet-optical elements, and, earlier this year issued an RFI based on these requirements (dubbed long haul OTP). We note that, as in the case of Verizon’s original metro/regional packet-optical transport RFP, the operator is far from alone in its requirements. In short, core P-OT elements will have the following characteristics:
- MPLS-TP centric with ODU-2 and above support
- Lots of wavelengths with 8-degree WSS optical switching
- 10s of Tbit/s aggregate switching capacity (including OTN switching)
- 100 Gbit/s per channel transport required
— Sterling Perrin, Heavy Reading