Any Format Any Speed
Any Mix
Flexibility in a Dynamic World
Anyone who has ever bought a new computer knows the sinking feeling that even before you get it home your brand new computer will be obsolete. The situation in telecom today is almost as dynamic. No sooner have carriers started to deploy BPONs to deliver 20Mb/sec to the home customer than they are faced with the need to upgrade those customers to GPONs supporting 100Mb/sec. The ink isn’t even dry on the subscriber agreements. Further, even before they start the GPON deployments, they’re faced with the clamor for 1Gb/sec.
The situation in the core telecommunications network is hardly less dynamic. Back in the boom days, carriers were concerned with only a few formats and speeds, primarily OC-48 and OC-192 SONET/SDH with only a few of the more adventurous considering gigabit Ethernet (GE). Now we are faced with these plus OC-768, 10 and 40 GE. A growing number of carriers are now also managing “private wavelengths”, VPNs at layer 1. And this doesn’t even begin to address the issues of multiplying control planes.
Today’s plethora of line rates and protocols is nothing new, and it will only get more complex with time. In spite of all the proposals you hear for a unifying format or protocol, one ring to rule them all, in the real world it won’t happen. The networks are continually evolving. New speeds and formats will be added as they prove economic, and old pieces of equipment will stay in place until they are totally uneconomic, probably dead. For the most part, forklifting out old equipment just doesn’t occur in the major telecommunications networks. Consider that for years we have been hearing that the world was IP and that SONET was dead. Today, there are 400,000 SONET/SDH rings worldwide, a number that isn’t going to zero any time soon.
One result of this rapid network evolution is that we are increasingly seeing DWDM systems which must support a broad range of speeds and formats. Recently, an undersea cable operator reported to us that of the wavelengths on their fiber, 60% were SONET/SDH, 20% were IP, and 20% were “private”. How do you create, monitor and protect lightstreams in such a diverse environment?
While ROADMs significantly improve the situation in ring architectures, most of the “optical crossconnects” used to implement more sophisticated mesh architectures are actually electrical crossconnects with lasers. Over the last few years, the line cards for these electrical crossconnects have greatly improved their ability to handle multiple formats and speeds. However, there is always a limit to what they can support. Today that limit is typically 10 Gb/sec. If a carrier needs to add a few 40 Gb/sec lightstreams, then all the electrical equipment, transponders and crossconnects in the path must be replaced with new equipment. This includes not only the line cards but also the core electronic switching fabric of the crossconnects, impacting not only the new lightstream, but all lightstreams passing through that crossconnect. All of this translates into both large CAPEX investments and lengthy service activation times.
As with ROADMs, many of the great benefits of true photonic crossconnects comes from their ability to create, monitor and protect lightstreams at layer 1 without needing to look at the data bits in the traffic bearing channels. Thus, they not only eliminate all of the expensive, inflexible electronics, but are also immune to changes in format or speed.
With true photonic crossconnects, adjacent lightstreams may carry entirely different traffic formats and/or speeds. In addition, a single lightstream might carry 10 GE one second and 40 Gb/sec SONET traffic the next. This flexibility permits a single protection path to be shared among multiple traffic bearing paths, each carrying different formats and speeds. This is true N+1 protection, where the 1 does not need to know the nature of the traffic carried on the N.
Any Format, Any Speed, Any Mix.
